TC-BVAU-7412.pdf - Bibliothèque et Archives Canada

"THEY RECOGNIZE NO SUPERIOR CHIEF"

POWER, PRACTICE, ANARCHISM AND

WARFARE IN THE COAST SALISH PAST

by

WILLIAM O. ANGELBECK

M.A., The University of Missouri, 1997B.A., Southwest Missouri State University, 1992

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

in

THE FACULTY OF GRADUATE STUDIES

(Anthropology)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

April 2009

© William O. Angelbeck, 2009

ABSTRACT

This inquiry focuses on warfare in the Coast Salish past. Located in the

Northwest Coast of North America, the Coast Salish practiced warfare as a basic

component of their culture, and warfare manifested in two main periods.

Archaeologically, fortified defensive sites were constructed from 1600 to 500 BP.

According to ethnohistoric documents and oral histories, conflicts also erupted in the

decades after Euroamerican contact, about AD 1790. For this study, I incorporate

archaeological, ethnographic, ethnohistoric, and oral historical data for an investigation

of warfare, including Coast Salish practices, protocols, and ideology. I assess the types

of settings in which warfare occurred and evaluate the motivations for conflict. Finally, I

examine these practices for insights into Coast Salish sociopolitical organization and

how it altered through time.

To evaluate the array of data, I employ a theoretical framework integrating

power, practice, and anarchism. For power, I implement Eric Wolf’s modes of power to

assess the intensity of conflicts and scales of defensive site construction. For practice, I

harness Pierre Bourdieu’s materialist approach to culture, which is focused on historical,

human actions, or practices; moreover, Bourdieu’s multiple types of capital provide a

rubric for assessing motivations for warfare as individuals pursue and exchange various

forms of capital. The theory of anarchism provides principles for evaluating the

dynamics of societies without formal governments. These include an emphasis on local

autonomy, voluntary association, mutual aid, network organization, and the

decentralization of authority (and resistance to concentrations of authority). This

framework illuminates how these principles varied throughout the Coast Salish past and

highlights significant differences in defensive structures between precontact and

colonial periods.

–– ii ––

Both periods of warfare appear after phases of increasing entrenchment of elite

power and hegemony (2400 - 1600 BP and ca. 500 to 200 BP). Both periods also exhibit a

broader expanse of elites, or nouveau riche. I conclude that warfare was an anarchic

practice implemented by Coast Salish factions to destabilize elite power structures and

allow non-elites to gain wealth and prestige. These practices resulted in the

decentralization of power––a heterarchy of chiefs, rather than a centralized chiefdom.

–– iii ––

TABLE OF CONTENTS

Abstract ii..........................................................................................................................Table of Contents iv........................................................................................................................List of Tables vii.......................................................................................................................List of Figures viii......................................................................................................................Acknowledgments x.........................................................................................................................

Chapter I: Introduction 1................................................................................................................The Subject of Warfare 4.....................................................................................................Nested Levels of Analysis: Power, Practice, and Anarchism 6....................................The Theoretical Approach and Argument 7....................................................................

Chapter II: The Theoretical Framework––Power, Practice, and Anarchism 15...................Power 15........................................................................................................................Practice Theory 22...............................................................................................................Anarchism 27........................................................................................................................

Core Principles of Anarchism 28................................................................................Individual and local autonomy 29.......................................................................Voluntary association 29........................................................................................Mutual aid 30...........................................................................................................Network organization 32.......................................................................................Decentralization and antiauthoritarianism 34....................................................Justified authorities 35............................................................................................

An Anarchist Theory of History 36............................................................................Integrating Anarchism, Practice, and Power 39..............................................................

Chapter III: The Anthropology and Archaeology of Warfare 40............................................The Concern with Origins 40.............................................................................................

Scalar Approach to Power 43......................................................................................The Concern for Causes 49.................................................................................................Anarchist Theories of Warfare 53......................................................................................

Interrelating Scales through Alliance Formation 58................................................Conclusion 65.......................................................................................................................

Chapter IV: Histories of Warfare from Documents and Oral Histories 69...........................Early Accounts of Warfare 69.............................................................................................

The Fort Langley Journals 73.......................................................................................Accounts from the 1840s and 1850s 75.......................................................................The Memoirs of Samuel Hancock 77..........................................................................

A Scalar Approach to Accounts of Warfare 79................................................................Expressions of Increasing Modes of Power 79..........................................................

(i) Power as an attribute of a person, or individual power 80..........................(ii) The ability of one to impose its will on another, or relational or

interpersonal power 80....................................................................................(iii) The control of social settings, or organizational power 81.......................(iv) The ability to establish or demolish the settings themselves, or

structural power. 83.......................................................................................A “Continual State of Fear”: The Power of the Lekwiltok 88........................................Colonist and Coast Salish Conflict 94...............................................................................

–– iv ––

Conclusion 96.......................................................................................................................

Chapter V: Welfare and Warfare: Modes of Production and Destruction 99.......................Weapons, or the Means of Destruction 101......................................................................

Melee Weapons 102......................................................................................................Projectile Weapons 103.................................................................................................

Equipment and Appurtenances for Warfare 105............................................................Armour 105....................................................................................................................War Canoes 106.............................................................................................................Defensive Architecture and Head Poles 108.............................................................

Warriors 109......................................................................................................................Leaders for Defense 114...............................................................................................Roles for Defense 116....................................................................................................Women and Warfare 117.............................................................................................Organization of Defense 118.......................................................................................Warriors and Fighters 119............................................................................................Leaders for Offense 121................................................................................................

Protocols of Conflict 122.....................................................................................................The Forces of Destruction 127............................................................................................

The Various Forms of Capital 129..............................................................................

Chapter VI: An Archaeological History of Warfare in the Northwest Coast 137................Initial Colonialization & Archaic Period (ca. 11,000 to 3500 BP) 138...........................Early Pacific (4500 to 3500 BP) 143.....................................................................................Middle Pacific (3500 to 1500 BP) 147.................................................................................

Early Middle Pacific (3500 to 2000 BP) 147................................................................Late Middle Pacific (2000 to 1500 BP) 152.................................................................

On the development of an elite class 155............................................................On storage and its implications 157.....................................................................

Late Pacific (1500 BP to contact) 159..................................................................................Postcontact Period 163.........................................................................................................Conclusion 167.....................................................................................................................

Chapter VII: Lookouts, Refuges, Fortifications, and Stockades 168......................................The Case for Defensiveness 168.........................................................................................Defensive Aspects of Residential Villages 170................................................................

Plankhouses 170............................................................................................................Household Arrangement 173......................................................................................

Defensive Sites 174...............................................................................................................Lookouts 174..................................................................................................................Refuges 180....................................................................................................................

Blockhouses 183......................................................................................................Underground houses 185.......................................................................................

Trench-Embankment Fortifications 190.....................................................................Bluff top fortifications 195.....................................................................................

Indian Fort Site (DgRr-5) 195..........................................................................Cardale Point (DgRv-1) 196............................................................................

Peninsular spits 201................................................................................................Rebecca Spit (EaSh-6) 201................................................................................Manson’s Landing (EaSf-1) 204......................................................................

Rocky headlands 204..............................................................................................EaSd-3 205..........................................................................................................Manor Point (DbRv-13) 207............................................................................

Rock-Wall Fortifications 209........................................................................................Stockades 213.................................................................................................................

–– v ––

Labour Organization 215....................................................................................................Conclusion 218.....................................................................................................................

Chapter VIII: Defending Against Whom? 222..........................................................................Coast Salish Sociopolitical Organization and Defense 226............................................Internal versus External Warfare 227................................................................................The Battle at Maple Bay 229...............................................................................................The Contextual Nature of Enmity and Alliance 241.......................................................

Chapter IX: Autonomy and Alliance 244.....................................................................................The Distribution of Defensive Sites 245............................................................................Archaeological Evidence for Networks of Defense 248.................................................Historic Evidence for Strategies of Defensive Site Distribution and

Cooperation 257.............................................................................................................Indications for the Increasing Frequency of Attacks 261...............................................Conclusion 275.....................................................................................................................

Chapter X: Elites, Hereditary Tradition, and Limitations to Social Mobility 276...............New Forms of Capital: The Development of the Nouveau Riche 277...........................The Balkanization of Marpole 281.....................................................................................Short-Cutting Traditional Practices to Higher Status 283..............................................Limiting or Enhancing Social Mobility––The Poles of Entrenchment and

Flexibility 289.................................................................................................................The Centrifugal Nature of Coast Salish Warfare 296......................................................Oral Histories and Accounts Concerning the Abuse of Power 301..............................

Chapter XI: Conclusion 304............................................................................................................Summary of Inquiry and Arguments 304........................................................................Suttles’ Quandaries 312.......................................................................................................

Bibliography 315............................................................................................................................

Appendices 354............................................................................................................................Appendix A: Sources for Defensive Sites Listed from North to South from Archaeological, Ethnohistoric and Ethnographic Sources. 354....................................

–– vi ––

LIST OF TABLES

Table 1: Types of coalitions (adapted from Tattersall 2006) 63..................................................Table 2: General chronology for the Northwest Coast and Gulf of Georgia. 139..................Table 3: Size of Smelt Bay village, its defensive houses, and regional defensive sites. 253...Table 4: Radiocarbon dates recovered at trench-embankment sites. 262................................Table 5: Historical and ethnographic records documenting dates for stockades. 263...........Table 6: Size of stockades from ethnohistoric descriptions. 264................................................Table 7: Underground house size and depth. 264........................................................................Table 8: Trench-embankment fortification sizes. 265.................................................................Table 9: Average size of underground houses and trench-embankment sites to

postcontact stockades. 266..................................................................................................Table 10: Maximum midden depths within the defended areas of trench-

embankment sites. 270........................................................................................................Table 11: Traits of Late Period trench-embankments versus postcontact stockaded

villages. 271...........................................................................................................................

–– vii ––

LIST OF FIGURES

Figure 1: Map of Coast Salish area, including most groups mentioned in the text. 3............Figure 2: Map showing locales of early encounters, forts, and most places

mentioned in the chapter. 70.............................................................................................Figure 3: Map showing approximate spread of smallpox in 1782 and population

estimates in the 1830s. 92....................................................................................................Figure 4: Model of the exchangeability of various forms of capital for the Coast

Salish. 130......................................................................................................................Figure 5: Example of plankhouse as fully wooden enclosure; from a detail of a

photograph taken at a potlatch in Songhees territory in 1874. 171...............................Figure 6: Lookout sites in the Coast Salish area as noted in ethnographic and

archaeological sources. 175.................................................................................................Figure 7: Locations of refuge areas and blockhouses. 181........................................................Figure 8: Location and depiction of "Block House" at Lamalcha Bay village based

on British accounts in April of 1863. 184........................................................................Figure 9: Underground refuge locations in the Coast Salish area from

archaeological and ethnographic sources. 186................................................................Figure 10: Semisubterranean pit features at Penn Cove, Whidbey Island (45IS50),

interpreted as underground refuges by Bryan. 187........................................................Figure 11: Wireframe surface map of an underground house (UH 1) at Smelt Bay

(EaSf-2) with plankhouse outline to the north (left) to indicate surface level. 189...Figure 12: Newcombe’s drawing of two trench-embankment types on southern

Vancouver Island, including the peninsular bluff at Albert Head and a triple trench-embankment at Cadboro Bay. 191...................................................

Figure 13: Locations of trench-embankment fortifications. 193.................................................Figure 14: Buxton’s depiction of three major types of trench-embankments. 193..................Figure 15: Typology of trench-embankment sites for this investigation. 194..........................Figure 16: Harlan I. Smith’s photograph of DgRr-5 from August 7, 1915, with a

view of the trench. 195........................................................................................................Figure 17: Drawing and interpretation of Indian Fort Site, DgRr-5, by Don Welsh. 196.......Figure 18: Surface map of Cardale Point (DgRv-1). 197..............................................................Figure 19: View of eastern portion of trench embankment. 198..............................................Figure 20: View of embankment and trench from the east, Profile Trench 1. 199...................Figure 21: Profile of eastern portion of trench embankment at Cardale Point as

measured with a total station (Profile Trench 1). 200.....................................................Figure 22: Stratigraphic profile of the lower portion of eastern trench-embankment

(Trench Profile 1). 200........................................................................................................Figure 23: Surface map of Rebecca Spit (EaSh-6), reconstructed from Mitchell’s

contour map. 202.................................................................................................................Figure 24: Mitchell’s stratigraphic profile of the top portion of the western trench-

embankment feature, showing stakemould of barricade and midden material abruptly stopping at barricade wall. 203.........................................................

Figure 25: Surface map of Manson’s Landing trench-embankment site (EaSf-1). 205............Figure 26: Surface map of fortification in Desolation Sound (EaSd-3), view from

the east. 206......................................................................................................................Figure 27: Village near Bute Inlet with house-floors at top of slope extending out-

ward protectively; detail of a drawing sketched by a member of Vancouver’s crew in 1792. 207...........................................................................................

Figure 28: Surface map of Manor Point (DbRv-13). 208..............................................................Figure 29: View of trench at Manor Point (DbRv-13) from the east, from the

–– viii ––

highest point on the promontory behind the trench. 209.............................................Figure 30: Map of rock-wall fortifications in the Coast Salish area from

archaeological sources. 210.................................................................................................Figure 31: View of rock-wall feature at Xelhálh in the Fraser Canyon. 211...........................Figure 32: Network of rock-wall fortifications in the Fraser Canyon. 212................................Figure 33: Map of stockaded villages from ethnohistoric and ethnographic sources. 215....Figure 34: Core areas for three defensive practices and the peripheral extent for

the sharing of each practice. 220......................................................................................Figure 35: Trench-embankment sites recorded by Bryan and Buxton in relation to

boundaries of Wakashan groups. 223...............................................................................Figure 36: Defensive sites in the Coast Salish area. 225...............................................................Figure 37: Types of conflicts in compendiums of accounts of warfare, ethno-

historically at Fort Langley and ethnographically. 228................................................Figure 38: Map of the Battle of Maple Bay, indicating coordination of Coast Salish

groups and roles. 235...........................................................................................................Figure 39: Scalar portrayal of social organization and corresponding defensive

manifestation. 243................................................................................................................Figure 40: Surface map of Smelt Bay (EaSf-2), Cortes Island, indicating two

underground houses. 250...................................................................................................Figure 41: Late Period defensive sites in the Smelt Bay region. 251........................................Figure 42: Smelt Bay village site compared to average size of regional defensive site

areas. 254......................................................................................................................Figure 43: Northern Gulf Islands network of defensive sites. 255.............................................Figure 44: Map of the Battle at Lamalchi Bay, indicating coordination of defensive

efforts. 260......................................................................................................................Figure 45: Radiocarbon dates recovered at trench-embankment sites. 262..............................Figure 46: Chart comparing the average size of underground houses and trench-

embankment sites to postcontact stockades. 266............................................................Figure 47: Battle Between Clallam and Makah at I-eh-nus, 1847, by Paul Kane 268..............Figure 48: Drawing by Spanish artist of “Indian fortification on the Strait of Juan de

Fuca.” 268......................................................................................................................Figure 49: Elmendorf’s sequence of social class evolution among the

Twana. 284......................................................................................................................Figure 50: Number of radiocarbon-dated burials according to presence/absence of

cranial deformation. 293......................................................................................................Figure 51: Cranial deformation in burials by period, including those dated by

radiocarbon association, interpretation, or radiocarbon dates. 293..............................Figure 52: Chronological chart indicating periods of entrenchment followed by

periods of warfare. 298........................................................................................................

–– ix ––

ACKNOWLEDGMENTS

Many people helped me with this project in various capacities. Foremost, I want to acknowledge my committee, from whose scholarship I have learned a great deal––whose ideas have influenced my thought throughout this work. Having come from doing archaeology in the Southeast and Plains, I was initially unsettled by the lack of ceramics and the comparative paucity of lithics. R.G. Matson’s work and teachings on the Northwest Coast helped me find a footing in the region’s culture history and think through the theoretical issues at debate. Bruce Miller was also foundational in this respect, helping to situate my understanding of Coast Salish culture and its political context. My advisor, Michael Blake has really helped advance my thought, particularly concerning broader archaeological theory and the development of social inequality. I respect his emphasis on collaborating with local communities, situating the contemporary importance of archaeology. It was my interest in these subjects and practices that drew me to study in the Department of Anthropology at UBC. I gratefullyacknowledge the years of financial support through the Charles and Alice Borden Fellowship––support that has made this whole endeavor possible.

I have learned a great amount from my colleagues with whom I've been able to work over the past several years. Foremost among these are Colin Grier, Eric McLay, and David Hall. Colin Grier helped me to sharpen some of the ideas presented here as well as contributing to fieldwork at Cardale Point. Eric McLay enabled much of my work in Hul’qumi’num territory and donated his time to these efforts, for which I am indebted.

I tip my hat to many others who helped me with work in the region. Dave Schaepe provided many leads for my work in Stó:lō territory. Rudy Reimer provided a tour of defensive sites in Squamish territory for my archaeological field school and has also been generous with other information about the area. Darcy Mathews helped to provide access to one site for this fieldwork. He and Pete Dady were helpful with information about defensive and other sites on Southern Vancouver Island.

Others provided access to many reports and documents that have proven extremely valuable. Thanks are due to Grant Keddie, who provided me with his reportson defensive sites in the Victoria area and led me to several oral histories of warfare. I have also spoken with others about information, reports or accounts that have proven useful: Donald Mitchell, Richard Brolly, Al McMillan, Rob Field, Peter Merchant, SimonKaltenrieder, Al Mackie, Robert Mierendorf, Eric Forgeng, and Don Welsh.

At UBC, I want to acknowledge the many discussions that I have benefitted from, including those with Jesse Morin, Nadine Gray, Gaston Gordillo, Brian Chisholm, David Pokotylo, Patricia Ormerod, Sean Lauer, Sue Rowley, Charles Menzies, Kisha Supernant, Andrew Martindale, and Zhi-Chun Jing.

I thank my field school team for their work at several sites, including Sean Aldcroft, Amy Davidson, Erin Hannon, Marlowe Kennedy, Matt McGinity, and Stephanie Watts, as well as Mark Harry and Al Hanson from the Klahoose First Nation. Special thanks to Meng Ying who helped lead the field school and helped to orchestrate the mapping for some of the sites included here. I thank Steven Acheson at the B.C. Archaeology Branch for facilitating the permit for this work, and Chris Kissinger for helping with the permit to work in BC Parks.

It has been an honor to work in the Coast Salish region. I have gained many insights from interviews and discussions I have had with many people throughout the region. This is especially so for the weeks spent at Frank Malloway’s longhouse in Stó:lōterritory during the ethnographic field school. I want thank Sonny McHalsie for the tourof the area he provided and the discussions we’ve had. I want to thank the Klahoose

–– x ––

First Nation, particularly former Chief Duane Hanson and Kathy Francis. I also thank Ken Hanuse, who guided our team throughout Desolation Sound. At the Sliammon First Nation, I thank Chief Walter Paul, Maynard Harry, and Norman Gallagher. I am grateful to tour the Upper Skagit Tribe members for their hospitality and for providing me with a tour of their territory, especially Sherman Williams, Floyd Williams, Vi Hilbert, and Scott Schuyler. I thank Leona Sparrow for helping me to work in Musqueam territory, and I also greatly appreciate the interviews and site tours with Florence James of the Penelakut and Arvid Charlie while working in Cowichan territory.Their histories about the battles at Maple Bay and Lamalchi Bay have been especially valuable. Previously, I had conducted most of my archaeological work in Missouri where the Osage and Missouri peoples were no longer present in the state, having been forced to Oklahoma reservations. So, even though I hiked and surveyed most creeks and passes in the Ozarks, I never once heard a place-name. It has been powerful for me to hear Coast Salish place names and histories and to witness Coast Salish peoples’ strong connection to the land and sea. I support their continuing defense of their territory and intend to continue working with them in helping to preserve their rich heritage.

Lastly, I want to thank my family who have long supported my efforts, especially my father, along with Jonn, Debbie, and Mike.

To Kristin, sine qua non––without your support, partnership, and camaraderie this work could not have been completed. And, to my daughter, Simone, I will always be ready to stop working and play––that's been helpful, too. Let's go.

–– xi ––

Chapter I: Introduction

There are still many unanswered questions concerning the pre-contact culture of the Coast Salish of the Strait of Juan de Fuca and adjacent waters. Two of the most important of these have to do with authority and conflict.––Wayne Suttles (1989:251)

In July of 1741, the Russian ship, St. Paul, of the Bering expedition steered within

sight of Chicagof Island and sent a boat of ten well-armed men toward shore. After no

sign for seven days, they sent another boat. Then, no sign from either. The next day,

two Tlingit canoes approached their ship, and apparently with each viewing the other as

hostile, they did not make contact. In the ship’s journal, they recorded: “We then

became convinced that some misfortune had happened to our men” (Golder 1922:296).

The first European encounter with the Northwest Coast seemed to have resulted in a

violent fate. In 1792, Spanish explorers also soon became embroiled in conflict. After

the sudden killing of one of their officers on the Olympic Peninsula by unknown

perpetrators, they fired cannons at the next canoes they encountered, likely Makah,

possibly Klallam or Straits Salish (Whitlam 1989). A decade later, the crew of the Boston

would be massacred by the Nuu-chah-nulth, leaving only two survivors, captured as

slaves (Jewitt 1987 [1815]). Not all early contacts resulted in conflict, as many Northwest

Coast groups also were eager to trade for new kinds of goods, especially iron and

firearms (e.g., Gormly 1977; Gunther 1972), but these accounts indicate tensions and

conflict were commonplace.

Warfare was a ubiquitous part of life for the Northwest Coast peoples for more

than a millennium. The evidence for warfare is found in the weapons they made, the

armour worn, the villages that were fortified or camps hidden from plain view. Indeed,

the evidence for wounds and scars of violent trauma has been documented from bones

unearthed in burials. War also served as a path for achieving success and acquiring

–– 1 ––

status. Through the bounty of war, one could acquire loot and supplies to hold a

potlatch ceremony, or, more permanently, to control a productive salmon stream and its

bounty thereafter. Warfare was a way to avenge any slights to one’s character as well.

In this manner, the cycle of warfare was embedded into cultural practices of the

Northwest Coast.

It is unknown how long ago warfare occurred in the region, but defensive sites

that are archaeologically visible began to appear throughout the Northwest Coast by

about 1600 years ago, and began to proliferate around 1100 to 600 years ago (Moss and

Erlandson 1992). Another period of defensive site construction occurs shortly after

Euroamerican contact, and I demonstrate that this was no mere coincidence.

In this work, I analyze the archaeological evidence for warfare in the Coast Salish

region (Figure 1), which has received less attention about warfare compared with the

northern Northwest Coast. In general discussions of Northwest Coast warfare,

examples are prone to highlight the warriors of the Haida, Tsimshian, Tlingit, or––

perhaps the most renown in historical memory, due to the Fort Langley journals

(Maclachlan 1998)––the Kwakwaka’wakw1 Lekwiltok, who menaced those to the north

and south of them from their bases in the Johnstone Strait and Discovery Passage.

In presenting this history, I begin by establishing the theoretical approach I

employ to interpret the archaeological evidence. This study is anthropological as well as

historical, and so I situate the archaeology of Coast Salish warfare within ethnohistoric,

ethnographic, and oral historical knowledge. The historical and ethnographic evidence

is better suited to understanding warfare of the late precontact to postcontact periods,

however, these sources do provide significant insights into warfare in the more distant

period, however, these sources do provide significant insights into one of the two

known efflorescences of warfare in the more distant Coast Salish past. An historically

and ethnographically informed assessment of the postcontact rise of warfare

1. Formerly, the Kwakwaka’wakw were known as the Kwakiutl (e.g., Suttles 1990c).

–– 2 ––

-128 

-128 

-126 

-126 

-124 

-124 

-122  

-122 

-120  

-120 

46  46 

48  48 

50  50 

52  52  

54  54  

0 50 100 km

-126 -124

50

49

48

BRITISH COLUMBIA, CANADA

WASHINGTON, UNITED STATES

Twana

SOUTHERN LUSHOOTSEED

NORTHERN LUSHOOTSEED

Quinault

LowerChehalis

UpperChehalis

Cowlitz

Klallam

HUL’QUMI’NUM

HALQ'EMÉYLEM

HUNQUMINUM

Squamish

Nooksack

Homalco

Klahoose

SliammonSechelt

Pentlatch

Comox

Nuxalk

NORTHERN COAST SALISH

CENTRAL COAST SALISH

SOUTHERN COASTSALISH

SOUTHWESTERNCOAST SALISH

Stó:lo

UpperSkagit

LowerSkagit

Duwamish

Snohomish

Suquamish

Cowichan

LANGUAGE GROUPSALISH REGIONS

Coast Salish Group

SookeSonghees

Skykomish

Snoqualmie

Puyallup

Musqueam

SemiahmooSnuneymuxw

Lummi Saanich

Nisqually

Fraser River

Skagit R

iver

Columbia River

PACIFIC OCEAN

Juan de Fuca Strait

Olympic Peninsula

Vancouver Island

Gulf of Georgia

Puge

t Sou

nd

NUU-CHAH-NULTH

MAK

AH

KWAKWAKA’WAKW

Ooweekeeno

QUILEUTE

CHINOOK

Tillamook

Heilsuk

Haihais

Tsimshian

Carrier

CHILCOTIN

LILLOOET

NLAKA’PAMUX

KLIKITAT

YAKIMA

Umatilla

ADJACENT CULTURAL GROUPS

Seattle

Portland

Vancouver

Victoria

(CHEMAKUM)

City

Figure 1: Map of Coast Salish area, including most groups mentioned in the text.(Locations are approximate [after Suttles 1990a; 1951; Barnett 1955; Haeberlin and Gunther 1930.)

–– 3 ––

is used to assess, interpret, and contrast with the archaeological evidence of the

precontact period of warfare, in an application of the direct historical approach (e.g.,

Marcus and Flannery 1994; Wedel 1938). In presenting the archaeological evidence, I

discuss the range of these defensive site types, their associated technologies (both tools

and features), and I make inferences about the strategies and tactics associated with their

construction. I also discuss how warfare and defensive sites provide insights into how

Coast Salish peoples imbued social relations with power; how they conducted, altered,

and shared their practices; and how they handled and arranged their sociopolitical

organization. Before I summarize the overall approach, I discuss why the topic of

warfare provides a useful focus for such a study.

The Subject of Warfare

Warfare is a subject that has been studied in many ways and from multiple

vantage points. Conflict, as journalists know, has inherent drama and so they use those

tensions––whether violent, sporting, or political––as the focal points of their stories.

Conflicts help to clarify issues and events, and an audience is often drawn to such

narratives. The catastrophic and eruptive nature of battle and clear opponents are

ready-made for a news story, an advantage over developments that occur at a gradual

pace, although perhaps no less different in result. With the archaeology of warfare, the

specific stories of conflict are largely gone, although its analysis can still highlight the

broader tensions and conflicts in the past.

Much of recent archaeology, since postmodernism, has emphasized agency (e.g.,

Robb and Dobres 2000), and I cannot think of anything more agential than the defense of

one’s community. Often postprocessual archaeologists discuss agency through various

symbolic expressions and rituals––there has even been a discussion of agency as

expressed through postmoulds (e.g., Pauketat and Alt 2005). However, agency is at its

most concrete in its assertion or defense of physical control, the taking and protection of

goods, resources, or territory; in causing the death of others to protect or enhance one’s –– 4 ––

own life and status. Here, warfare with its active expressions can be used to interpret

the archaeological record in more humanistic terms, not simply as resource gatherers or

as markers for a cultural-typological signature, but as those fulfilling lives and having

traditions worth fighting for. An analysis of warfare is one way of seeing the past in a

more active, dynamic light that is more expressive of the behaviors of past peoples.

Depending on general cultural values, a warrior’s success in battle may be

highly valued by his community and his successes glorified, bringing him both status

and wealth. However, in some societies––as among some Coast Salish groups––warfare

may be regarded negatively, particularly when not carried out in defense. Similarly,

many acts initiating warfare in recent history and in the past have rationales for being

defensive actions.2 These ideological explanations imply an awareness that offensive

actions may not be seen as just. For this work, however, warfare is simply another

practice engaged in by people. For an archaeology of warfare, it suffices that warfare

occurred in the past of the Northwest Coast, despite arguments in favour of or opposed

to war. Indeed, the reasons and rationales––pro and con––for warfare were as

multifaceted throughout the past as they are today.

Warfare is important for archaeologists because it contributes to structural

change through time, highlighting shifts in the parameters of a group’s social and

political operation. Settlement patterns of sites expand to include fortified sites or

refuges, and residential villages and camps may be moved to less accessible or visible

areas. A village burned in an attack might indicate its time of abandonment, or groups

may abandon a region during times of increased warfare. Groups also might take over

new territory. All of these actions might reveal shifts in a region’s culture history,

indicating structural change.

2. Examples include the German invasion of Poland that started World War II, which Hitler argued was a defensive act, merely “returning fire” after German agents (acting as Poles) staged an attack on a German radio station in Gleiwitz, Poland (Baker 2008:132-136). Even recently, the U.S. invasion of Iraq was purported to be “preemptive” against Saddam Hussein’s stockpiling of “weapons of mass destruction.”

–– 5 ––

Nested Levels of Analysis: Power, Practice, and Anarchism

In order to address broader aspects of meaning and rationale in warfare, I will

follow Trigger’s (1989, 1991) holistic archaeology, as others have done in the region (e.g.,

McMillan 1999). To better understand the archaeological record, Trigger (1989:235)

advocated the incorporation of ethnography, ethnohistory, linguistics, art history, oral

traditions, plus any other relevant sources. He found that Marxism provided a context

that “encourages the analysis of behavioral phenomena in as holistic a context as

possible” (Trigger 1989:235) because it provides a theoretical framework that integrates

economy and sociopolitical organization and allows for the interpretation of the broader

sociopolitical context through physical archaeological remains that are largely indicative

of economy. While this study will be holistic in Trigger’s sense, it also applies, more

specifically, three predominant approaches to warfare in the past through nested levels

of analysis involving power, practice, and anarchism.

As Trigger pressed for the utility of Marxism, I draw on Wolf’s (1990) modes of

power. Like Trigger, Wolf (1999:14-15) was a proponent of a Marxist-based or Marxian

approach, however, his conceptions of power are much more nuanced than Marx and

Engels’ version. Wolf’s modes of power provide for understandings of the degree or

scale of power, which is useful for a study of warfare. His model goes beyond a simple

typology because he assesses the increasing concentrations and applications of power.

Also originally influenced by Marxism, Bourdieu’s (1977, 1990) practice theory

provides for the interpretation of physical archaeological remains as the patterns of

historical and cultural practices, as opposed to functionalist processes (Pauketaut 2001).

Accordingly, the detritus and features of the archaeological record result from past

traditional practices; the habitual nature of practices structure or pattern the artifacts

and features found archaeologically. In so doing, it provides a way to connect specific

archaeological manifestations in the archaeological record to cultural traditions that

change over time.

–– 6 ––

For the third part of the framework, I will use the theory of anarchism to

interpret social organization. Since most societies throughout human history had no

formal government, my premise is that the theory of anarchism provides principles for

understanding non-state forms of social organization, such as that of the Coast Salish.

In a study of warfare, understanding the dynamics of past social organization is

important––as Malinowski (1936:444) had discussed. For him, warfare was the “use of

organized force between two politically independent units.” The emphasis in his

definition is on the social organization of combat and how warfare indicates

sociopolitical autonomy between competing groups, either groups asserting control or

temporary dominance of others, or through pursuing independence, as in revolution or

civil war. Accordingly, when warfare is present, there is no overarching entity or polity

that controls affairs––the rules of dominance are precisely being worked out through the

conflict itself.

To summarize, Wolf’s modes of power theory is used to understand the intensity

and application of physical domination (or attempts at such). Practice theory is used to

understand the traditional practices of warfare as indicated in the archaeological record

that involve assertions of power. Finally, anarchism is used to assess how such practices

were organized in societies without formal government. I discuss each of these three

approaches in more detail in the next chapter. Below, I summarize the work as a whole.

The Theoretical Approach and Argument

What follows is an inquiry into the role of warfare in the Coast Salish past.

Predominantly, this inquiry is archaeological, concerning the establishment of defensive

sites. Defensive sites are a fruitful avenue for a perspective on warfare––these are the

architecture for warfare. Additionally, I supplement this with other aspects of warfare

as well, such as weaponry, strategy, and tactics. I combine this information with the

ethnographic and ethnohistoric detail and the oral histories about Coast Salish warfare.

Defensive sites are large-scale constructions that exhibit certain traits unique to –– 7 ––

the Coast Salish, yet these also reveal regional variability within the Coast Salish area

overall. Of course, there are also changes through time, marking the shifts in the

intensity and frequency of warfare. Such large-scale constructions are also indicators of

the nature of Coast Salish sociopolitical organization, indicating the cooperative

endeavors of households, villages, or regions for protection.

To undertake this inquiry, I conducted investigations at several defensive sites in

the Strait of Georgia, both in the northern and southern Gulf Islands, and sites in the

lower mainland and Vancouver Island.3 Most investigations were aimed towards

surface mapping of these defensive features and core sampling. To buttress and inform

the archaeological data, this investigation also consisted of significant investigation of

available ethnographic, ethnohistoric, and oral histories.4 I have conducted

ethnographic interviews relating to warfare, recording stories of the Cowichan warrior

Tzouhalem, the Battle at Lamalchi Bay, and the Battle at Maple Bay.5

Indeed, one of the main periods of warfare begins in the wake of contact, the

3. The majority of the reports for the investigations conducted as part of this inquiry are or will soon be part of the library of permit reports managed by the Archaeology Branch of British Columbia, within the Ministry of Tourism, Sports, and the Arts in Victoria. These will include reports for the Indian Fort Site (DgRr-5; Angelbeck 2006) and the sites investigated in the northern Gulf Islands and Desolation Sound which include Smelt Bay (EaSf-2), Manson’s Landing (EaSf-1), and EaSd-3 (Angelbeck 2008a). A report will also be available for the investigations at Cardale Point (DgRv-1), on Valdes Island (Angelbeck 2008b). The investigations at Manor Point (Angelbeck 2008c) will be on file at the library maintained by the Laboratory of Archaeology at the University of British Columbia, as will the aforementioned reports.

4. Several archives were consulted as part of this inquiry. These included the Special Collections at the University of British Columbia, including the papers of Homer Barnett (1930-1940), Charles Borden (1905-1978) and Wilson Duff (1960-1976). I also researched at the British Columbia Archives and library in Victoria. In Seattle, I accessed the Pacific Northwest Special Collections at the University of Washington, including the papers of Erna Gunther (1871-1981). Special attention was given to Wayne Suttles’ papers (1946-1986), particularly notebooks from fieldwork from 1948 to 1952. Other archives that were consulted included the Cortes Island Museum and Archives in Manson’s Landing and the archives of the Klahoose First Nation in Squirrel Cove, also on Cortes Island. Other archives maintained by First Nations that I was able to consult included the Stó:lō Nation Archives in Sardis, B.C., and the archives of the Hul’qumi’num Treaty Group in Ladysmith, B.C. I also consulted the library of archaeological permit reports maintained by the Archaeology Branch and searched site forms through their geographical database online.

5. I conducted ethnographic interviews in the Coast Salish area and worked on ethnographic projects in Stó:lō, Upper Skagit, and Hul’qumi’num territories that have proved useful to this study (Angelbeck 2003; Miller and Angelbeck 2006, 2008), particularly, interviews conducted with elders about the battles at Maple Bay and Lamalchi Bay, which were conducted at the

–– 8 ––

time covered explicitly by ethnohistory, ethnography, and even many of the oral

histories of warfare. This postcontact data provides observations and documents useful

for interpreting the archaeological data for warfare, which concerns much of the last

1600 years BP.6 In turn, the archaeological data allows us a perspective to evaluate the

changes that occurred in warfare in the postcontact period, as defensive practices

evolved to counter changes in the frequency and intensity of warfare and changes in

sociopolitical conditions.

Warfare concerns the application of power, not only in the physical acts of the

power displays and dances of a great warrior, or one group dominating another, but

also through the networks of alliances that are called upon for attack or defense. To

assess the various applications of warfare, I employ Wolf’s (1990) treatment of the four

modes of power, which involve individual, interpersonal, organizational, and structural

forms of power. More than a typology, these categories represent a scale of power of

increasing dimensions. These modes allow us to evaluate the intensification of warfare

in the past, through increasing scales of organization for both attack and defense.

While the politics and battles of warfare involve the dynamic interplay of power,

individuals implement these actions often through an array of traditional means and

options available to them. I deploy practice theory, as developed by Bourdieu (1977,

1990) and archaeologically adapted by Pauketat (2001). Traditional cultural practices

pattern the physical deposition of artifacts and structure the nature of archaeological

features. Moreover, practice theory is useful for understanding the strategies and tactics

implemented by individuals in their pursuit of status and the acquisition of various

forms of capital. The cultural practices for any group are influenced by how a group

organizes to implement those activities.

To assess how cultural practices are organized––more particularly for this study,

how Coast Salish organized to conduct their warfare practices and construct defensive

battle locations (Angelbeck and McLay 2008). 6. Radiocarbon years Before Present, or A.D. 1950.

–– 9 ––

fortifications––I employ a materialist analysis. For many culture areas, Marxist analyses

have been quite an effective form of analysis––a frame that was materialist, yet

contained a body of theory enabling interpretations of sociopolitical and ideological

social structures from the material patterns of largely techno-economic remains (e.g.,

McGuire 1992, 2008; Spriggs 1984a; Kohl 1984; Patterson 2003; Gilman 1981; Benn 1990).

Here, I employ a materialist analysis that also shares a long history: that of anarchism.

The premise is that the theory and principles of anarchism can be useful for

understanding anarchic societies, or societies without formal governments. While

Marxist theory is robust, it has been heavily oriented to state societies––discussions of

“pre-capitalist”societies (e.g., Marx 1965 [1857-1858]) imply a teleology for the capitalist

state, indicating that these anarchic societies are even defined by their lack of capitalism

or statehood. Non-state societies were not a focus for Marx and Engels, which partially

explains the need for theoretical reworkings of Marxism for anthropologists (e.g.,

Meillassoux 1980; Wolf 1982; Bloch 1983).

Anarchism, on the other hand, is about small-scale social organization, societies

that form from the bottom-up, rather than those that are directed centrally from above.

Local organization, however, can lead to the operation and maintenance of larger

projects and even industrial endeavors as groups cooperate and federate into larger

scales of organization, albeit the locus the control remains on the local level. Contrary to

its common connotations of chaos that are inherent in “anarchy,” anarchism is about a

form of social order. Anarchism consists of a body of theory as old, if not older, than

Marxism, depending on when one believes it developed (Woodcock 1962; Marshall 1993;

Guérin 1970). It similarly shows a long debate and dialogue, both with Marxists and

internally amongst anarchist thinkers, that has sharpened its framework and resulted in

numerous strains of thought.

The core principles of anarchism include: local and individual autonomy and

expression, voluntary association, mutual aid, network forms of social organization,

decentralization, antiauthoritarianism, gift economies, and direct democracy. Rather

–– 10 ––

than providing a model formula for how societies should function, anarchist thinkers

emphasize core principles and practices that should be adapted to local settings and

historical circumstances. These practices would not be static but would be constantly

renegotiated to continue to adapt to the contingencies of new historical situations.

To many anarchists, these principles were not conceived during the 19th century

by theorists, but really are reflections of innate tendencies that will surface through

interactions of humans as social beings, principles whose options will occur as

individuals engage with others, not only as kin but as a part of a community. These

basic principles are coordinated in different fashions and in different degrees according

to the needs of groups on local scales. These dynamics can link and federate with other

social groups to operate and enact goals on larger scales, including urban and industrial

societies. This theoretical framework––of power, practice, and anarchism––is erected in

the following chapter, Chapter II. Subsequent chapters illustrate how these interlocked

theoretical perspectives can be implemented.

In Chapter III, I discuss the anthropology and archaeology of warfare, primarily

to illustrate how the theoretical framework of power, practice, and anarchism can

effectively buttress as well as address some weaknesses in prior approaches. I argue

that the utility of Wolf’s (1990) scales of power provides a method of evaluation that

matches and builds upon the work of Otterbein (2004) and Kelly (2000). Another focus

in the anthropology of warfare concerns the causes of warfare. I consider how practice

theory provides a framework that does not reduce the human complexity in the multiple

reasons for warfare to a single cause. Instead, by focusing on the exchangeability of

capital, I argue that reasons for warfare can be more readily encompassed with

subsequent transactions or exchanges from resources acquired in battle as well as

provide responses that coincide with human reasons for warfare. Finally, I argue that

anarchism adds to the discussion of the anthropology of warfare by stressing the

sociopolitical dynamics of the bottom-up organization in small-scale and anarchic

societies, which can implement larger scales of organization through the networks of

–– 11 ––

alliances. The theory of anarchism emphasizes that there are always forces within social

organizations that aim to inhibit the concentration and centralization of power.

The subject of Chapter IV concerns the ethnohistory of warfare in the Coast

Salish area since 1790, beginning with the first documents by the Spanish and continuing

through the postcontact period, a period wherein warfare intensified until about 1870. I

use Wolf’s (1990) scales of power to assess the different types of warring interaction. I

demonstrate that the full range of Wolf’s scales of power are represented, from

individual to structural power and argue that scales of warfare intensified at various

points over the last two millennia. I close with a discussion of how the Euroamerican

settlers themselves employed structural power, both in domination of Coast Salish and

other Northwest Coast groups.

I provide an ethnographic overview of Coast Salish practices for warfare in

Chapter V. I discuss the role of the warrior and the dangerous and unpredictable nature

of warrior spirit powers. I show that the temporary authority given to the warrior in

times of war indicates a principle for validating or justifying the nature of an

individual’s authority among the Coast Salish. From the discussion of the variety of

defensive practices, I show that the Coast Salish exhibit distinct practices among the

Northwest Coast groups, commonly using tactics that allow for more flexibility, both at

individual and at collective scales. I also assess the causes of warfare in the oral histories

and apply the perspective of practice theory, with its emphases on improvisation of

strategy and tactics within an array of cultural practices or structures. I argue here that

the exchangeability of capital reveals how the Coast Salish could employ manifold

rationales for warfare. Perspectives that try to reduce these reasons not only limit their

applicability but are also not consistent with Coast Salish motives for warfare.

For Chapter VI, I present an overview of the archaeology of warfare on the

Northwest Coast through the perspective of the theoretical framework of power,

practice, and anarchism. I posit that over the course of millennia the nature of power

altered through time, placing additional constraints on the freedom of an individual or

–– 12 ––

local group. I also consider how this changing nature of power is furthered or inhibited

by changes in the human population. I argue that these are not meant as determining

human cultural traits, but rather as settings that enabled actors to enhance their power

sociopolitically.

In Chapter VII, I assess the array of defensive sites from predominantly

archaeological sources, but also types described ethnohistorically and through oral

histories. One key point of this chapter is that there were defensive practices that were

unique to the Coast Salish. However, there was still a strong sense of local or regional

preferences for particular practices. In fact, I show that no one Coast Salish defensive

practice was employed by the Coast Salish as a whole. Rather, there were numerous

core centers for certain practices with peripheries of influence. I argue that these

patterns are reflective of alliance and interaction networks that Suttles (1987 [1960]) and

others have documented.

Next, in Chapter VIII, I analyze arguments for whom the defensive sites are

meant to defend against. Often the ethnographic and archaeological literature stresses

the battles of the Coast Salish against western and northern groups, such as the Nuu-

chah-nulth and Kwakwaka’wakw. I argue that this notion needs to be reconsidered and

draw upon the oral histories and ethnohistoric accounts of warfare to assess the degree

of warfare that was common among the Coast Salish themselves. After considering

internal Coast Salish warfare, I also evaluate those battles between the Coast Salish and

external groups. I provide an account and analysis of the Battle at Maple Bay to show

how Coast Salish groups could freeze local tensions to ally in the face of larger threats, in

an example of bottom-up organization through networks of alliances. A main point is

that understanding the dynamics of their anarchic sociopolitical organization is critical

for assessing internal conflicts among Coast Salish groups and their corresponding

ability to unite into larger networks of cooperation and alliance against external groups.

I assess the evolution of defensive sites, in Chapter IX, to consider how some

defensive sites are meant for households while others are designed for allied households

–– 13 ––

or villages as a whole. Furthermore, I argue that defensive sites need to be considered

beyond the context of any particular site. Because of the nature of Coast Salish alliance

networks, defensive sites must be situated within regional contexts. Coast Salish areas

exhibit a distributed nature of defense that shows a flexibility of response at multiple

scales. I argue that Coast Salish defensive sites exhibit a range of possible defenses that

operate to the scales of the threats they faced, from attacks on particular houses to

villages or regions. Lastly, I demonstrate that the organization of of defensive sites over

the last 1600 years intensified to match the scale and frequency of warfare through time.

In Chapter X, I compare the two effloresences of warfare, the Late Period and the

postcontact period through a consideration of changes in sociopolitical organization. I

argue that both periods exhibit the growth of nouveau riche. In both cases, I argue,

warfare played a role in enabling commoners to gain wealth and status. Having been

blocked or inhibited in their ability to gain wealth and status through productive

methods, these individuals turned to warfare, or destructive methods, to make such

gains. In both periods, I argue that warfare served to restrict the concentration of power

among elites and redistribute it more equitably, if not in egalitarian fashion. Warfare

consisted of practices that enabled individuals and households to enhance their power

and autonomy. Viewed through the perspective of anarchist theory, the concentration

of power among elites came to be seen as entrenched and unjustified. Those blocked

from avenues to higher social status sought to disrupt the status quo and decentralize

the existing power structure.

In the conclusion, Chapter XI, I provide a summary assessment and revisit a

couple of Suttles’ quandaries regarding the role of authority and conflict among the

Coast Salish in light of this inquiry into the nature of warfare.

–– 14 ––

Chapter II: The Theoretical Framework––Power, Practice, and Anarchism

To undertake an analysis of warfare in the Coast Salish region, I use three

primary theoretical tools to assess the scale of power, the types of practices employed,

and the nature of their heterarchical social organization. For power, Eric Wolf (1990)

provides a treatment of four modes of power that increase in scale and effectiveness.

For practice, Bourdieu (1977, 1990) developed a theory that integrates structure and

agency, where long-standing traditions serve as an array of options for ready practices

that individuals can use as they compete with other individuals for various forms of

capital. For social organization, a theory is required that can encompass the fluid and

heterarchical nature of the Coast Salish, which is neither centrally hierarchical nor

egalitarian. While there were chiefs, there were not chiefdoms. There were elites,

commoners, and slaves without the centralization of stratified states. The theory of

anarchism provides principles for assessing such a society that does not fit general

anthropological models of social evolution based teleologically toward the centralization

of chiefdoms and states. Moreover, anarchism integrates well with both theories of

practice and power. Anarchism also provides a theory of history that incorporates

power, a point that anarchist theorists have repeatedly raised as a weakness in Marxism.

Anarchism also provides principles of social organization that assess how practices are

organized and implemented to ensure that individuals and local groups retain a high

degree of autonomy.

Power

Warfare and violence are expressions of power; in fact, these represent the

physical exertion of power of one group over another, or one individual over another.

–– 15 ––

In this sense warfare is a medium of social interaction where social and political power

is played out in the lives and histories of both individuals and groups. Eric Wolf (1990)

presented a framework for an analysis of power that provides valuable insight into the

nature of social relations that manifest in times of warfare and conflict. He conceived of

four “modes of power” (Wolf 1990:586-587):

(i) power as an attribute of a person, or individual power;

(ii) the ability of one to impose its will on another, or relational or interpersonal power;

(iii) the ability to influence or control individuals within social settings, or organizational power,

(iv) the ability to establish or demolish the settings themselves, or structural power.

His model is scalar in that each mode encapsulates the prior as nested levels or

dimensions of power, from the individual (personal) to self/other relationships

(interpersonal) to group dynamics (cultural) and structural governing (societal).

Violence and warfare can be expressed as physical manifestations of power accordingly,

from the individual to higher scales, which involve increasingly complex social

relationships. The first mode, individual power (i),7 is the personal power which is drawn

upon or displayed for purposes of politics and/or conflict. This is power as “potency or

capability, the basic Nietzschean idea of power” (Wolf 1990:586). These are the

characteristics or skills that may lead others to call on a particular individual to address

a need or resolve a problem. This is power as intrinsic to a person but does not concern

how that power is applied to others. The second mode, interpersonal power (ii), can be

expressed as power enacted between two individuals, as between a leader and

supporter, master and slave, or two combatants, as on a battlefield. Wolf (1990:586)

described interpersonal power as that “the ability of an ego to impose its will on an alter,

7. Throughout the text, I continue to refer to these four modes of power parenthetically in this manner.

–– 16 ––

in social action,” which is the interplay of one’s power in contest or conflict with

another. This mode describes the specific one-on-one interaction, but does not address

“the nature of the arena in which the interactions go forward” (Wolf 1990:586), or what

may be termed the social field according to Bourdieu (1977, 1990). The third mode of

power, organizational power (iii), can reflect power expressions within a group, as

through organization of others for defense or offense, or by factional competition for

positions of control. Wolf (1990:586) described this as the ability of one actor to

“circumscribe the actions of others within determinate settings,” which he also

characterized as a form of “tactical power.” Lastly, the highest form of power, structural

power (iv), is the ability to control or alter the social settings themselves, and this power

involves more than organizing others within an existing social arena; as Wolf (1990:587)

put it: “Structural power shapes the social field of action so as to render some kinds of

behavior possible, while making others less possible or impossible” (Wolf 1990:587).

[T]his is the kind of capital to harness and allocate labor power, and it forms the background of Michel Foucault’s notion of power as the ability “to structure the possible field of action of others.” Foucault (1984:428) called this “to govern,” in the 16th-century sense of governance, an exercise of “action upon action” (1984:427-428). Foucault himself was primarily interested in this as the power to govern consciousness, but I want to use it as power that structures the political economy. I will refer to this kind of power as structural power. This term rephrases the older notion of “the social relations of production,” and is intended to emphasize power to deploy and allocate social labor. These governing relations do not come into view when you think of power primarily in interactional terms (Wolf 1990:586-87).

Here Wolf wanted to concretize Foucault’s use of power from governance of

consciousness to the more material aspects of economy, after Marx. Marxists had

greatly improved understandings of power into realms involving governance and

ideology and control of labour and economy. However, just as Wolf wanted to reorient

Foucault’s emphasis, I would like to redirect this mode of power (as well as the other

modes) to material forces beyond economy, towards physical enactments of power:

warfare is the arena in which power unmasks itself for what it really is. Other more

hegemonic forms are understood to be backed by physical power, but in warfare, the

–– 17 ––

interaction is reduced to the fist, club, or cannon. Of the four modes of power, the

ability to alter the settings through structural power is the most destructive mode of

power. In applying this mode to warfare, it is necessary to conceive of it beyond simply

to “alter” or “orchestrate” the settings, as Wolf (1990:586-87) noted, but it also needs to

include the ability to destroy the settings themselves. In fact, this applies to all the

modes of power that Wolf defined––power indeed is social, economic, and institutional,

but it must have recourse to and foundation in its physical expression, as demonstrated

in warfare.

The differences between these modes of power and their intrinsic escalation in

dimension from one mode to the next can perhaps be readily illustrated by abstracting

these principles through chess. Wolf’s first mode of power, as intrinsic to the individual

or person (i), can be seen as the property of the piece concerned, whether a bishop that

moves diagonally, a rook vertically and horizontally, or the simple one-move advances

of a pawn. Those are the powers intrinsic to the individual piece regardless of their

relation to others. For the second mode of power, interpersonal power (ii), one piece can

take another (or has power over another) by virtue of their individual powers (i). A

rook, for instance, can take a pawn on its rank; a bishop can capture a knight along its

diagonal. The application of this power to an opponent’s king is noted in the term,

“Check,” meaning that power of one’s piece is threatening the opponent’s king. With

the third form, organizational power (iii) is applied through the coordinated

mobilization of one’s pieces, whether it is the combined attack of a pair of bishops or a

cordon of pawns; notably, to mate a king requires this organizational power in order to

win, as even the all-powerful queen cannot checkmate another king on her own (that is,

if the King is not inhibited in his movement or powers by his own defensive pieces).

The final form of power, structural power (iv), occurs when the organizational power of

one’s pieces orchestrates a scenario that controls the setting of the game. It is that point

when the setting is so structured that the opponent’s king cannot even make a move––in

other words: “Checkmate.”

–– 18 ––

However, as an abstraction, using the game has its limits. It does not illustrate

the obverse or destructive aspect of structural power which involves not controlling but

destroying the settings themselves, going beyond the boundaries of the board’s limits:

changing the rules mid-game––or swiping a backhand, clearing the pieces off the board

itself.

Of the four modes of power, structural power (iv) should be of interest to

archaeologists in studying warfare because it indicates points at which the conditions

can change, and change rather quickly. Warfare does restructure societal settings.

When novelists like Philip K. Dick (1962), with The Man in the High Castle, imagine a

North America in which Germany won World War II––a common type of theme in

imaginative fiction––these authors are playing on this aspect of the structural power of

warfare which change the social settings and the course of history itself.

I propose that these four modes of power, identified by Wolf, can readily be

identified as archaeological correlates in the Northwest Coast.8 Briefly, the burials of

warriors accompanied by their clubs, indicate individual power (i), the power of

individual warriors showcased with their weapons––turned from primarily functional

weapons into symbols of the warrior. For interpersonal expressions (ii), examples have

been well-documented by Cybulski (1992, 1994, 1999), among others, who have

documented trauma resulting from interpersonal violence. These include bone fractures

that are unlikely to result from falling out of a canoe: parry fractures on one’s forearms,

projectile points embedded in bone, beheadings, and fatal club imprints on skulls. Also,

the taking of slaves or the conversion of status from elite or commoner to slave is an

expression of such power.

Defensive sites such as fortifications or refuges, a focus of this study, mark an

example of organizational power (iii), as these sites involve a great amount of labour to

8. Kenneth Ames (1995:157) has also found Wolf’s (1990) modes of power useful for analyzing Northwest Coast societies, remarking that it is “a framework for understanding the power of coastal elite in the household and beyond the household.”

–– 19 ––

construct. Moreover, the death of a chief in battle would be a significant event, causing

political reorientation and renewed competition for power among and within the

household. A chief’s death might lead to a shift in organizational power, although the

societal settings and conditions would remain the same.

For structural power (iv), examples could be the demolishing of a village, or a

people. This also could be represented in the introduction of a new technology, such as

firearms, which might shift the settings and the balance of power advantageously

towards one group over another. Coast Salish groups demolished almost to extinction

the Chemakum people of the Port Townsend area, ca. 1845 to 1850 (Eells 1985:351; Boyd

1990:136; Elmendorf 1993:143-145). It is this latter level of warfare that causes

momentous shifts in the archaeological record, as one group conquers its neighbours

and spreads new practices and material symbols throughout the region—sometimes in

complete disregard of the subjugated peoples’ traditions or sites. Whereas

organizational power (iii) will have a pattern indicative of a sustained tradition of its

associated practices, such practices will generally leave a pattern of continuity.

Structural power (iv), on the other hand, is more likely to result in significant social and

material discontinuities.

As Wolf (1990) indicated, much of the recent interest in power in the social

sciences derives from the work of Foucault, who attempted to delineate the many

dimensions of power. In The Archaeology of Knowledge, Foucault (1972:7) stated that

“Traditional history attempts a coherent, continuous narrative, when history is marked

by discontinuities and irruptions.” Such discontinuities would result from power

exerted in structural ways, altering both settings and practices. Wolf’s discussion of

modes of power is useful for an archaeological analysis of warfare, because it provides a

scalar framework for understanding the modes of power expressed. Wolf (1990:587) did

not intend that these modes of power as typological categories, but rather as explanatory

devices for understanding human interaction: “it is the task of anthropology ... to

attempt explanation, and not merely description, descriptive integration, or

–– 20 ––

interpretation.”

Power, according to Flyvbjerg (2001), is not just an aspect that has been less

studied in anthropology, as Wolf had stated, or in the social sciences more generally.

Rather, Flyvbjerg (2001) considered analyses of power central to analyses in the social

sciences.9 After Foucault, Nietzsche, and Weber, he proposed that power should always

be analyzed as internal to all social relations, not as an external “tool” or other, but as the

medium of social relations. Moreover, power is not just a dominating, restrictive or

negative force, but is also productive and positive. Furthermore, it too often is focused

in individuals or centers when it should be analyzed as intrinsic to sets of relations

(Flyvbjerg 2001:131-132). The use of Wolf’s modes of power is a way to undertake a

phronetic approach to warfare in the past, situating power as a raw medium of

interaction. A phronetic approach, in its orientations to specific contexts and action, can

be aided archaeologically by a focus on practices, the patterned forms of actions.

9. His larger aim was to make the social sciences more relevant once again, since these disciplines have declined in societal importance in recent decades relative to the natural sciences. The reason, he argued, is that the social sciences mostly have tried to emulate the natural sciences, with their “physics envy,” producing universal laws, accumulating knowledge, and making predictions. Flyvbjerg (2001) argued that social sciences are an entirely different domain and that any effort towards approximating natural science is flawed. The answer is not to be in reducing social entities to physics or chemistry, rather it is deal with social science concerns through its own properties and dynamics. Part of the problem is that Western civilization emphasizes scientific knowledge and technology to the detriment of practical reason and ethics. Flyvbjerg discussed his three forms of knowledge as originally defined by Aristotle. Episteme is scientific knowledge (“epistemology”), based in analytical reasoning that generates universal principles that are invariable and context-independent. Techne refers to craft or art (“technology”/”technique”) and is pragmatic, variable, and context-dependent, although oriented towards production. The final form of knowledge is phronesis (which has no cognates in English), which refers to practical reason or ethics; it is pragmatic, variable, and context-dependent and is oriented towards action (Flyvbjerg 2001:57). The epitome of social science domain is the case study, which allows not for universals, but for examples of how humans relate. To Aristotle’s formulation, Flyvbjerg (2001) added that a conception of power must be integrated for a truly effective social science.

–– 21 ––

Practice Theory

All social life is essentially practical. All the mysteries which lead theory towards mysticism find their rational solution in human practice and in the comprehension of this practice.––Karl Marx (1970a [1845])

To study warfare from a primarily archaeological perspective, and with a focus

on the many modes of power associated with it, we must identify the practices of

warfare in the archaeological record. Practice theory, as articulated by Bourdieu (1977,

1990), permits an analysis of the basic practices that actively created the archaeological

record. In later periods these practices are also noted in ethnohistorical records,

ethnographies, and oral histories. For archaeologists, past practices (in Bourdieu’s

terms) contribute to the production of the archaeological record. That is, they extend––

in a rather simple way––beyond the descriptive nature of the artifact or feature to the

social traditions that produced those patterned actions. Almost fifty years ago,

processual archaeologists such as Lewis Binford (1962) exhorted archaeologists to do just

that as well. Binford criticized the established archaeological practice of building culture

histories as a particularistic, descriptive, and typological exercise. He and his students

advocated an approach that emphasized understanding the social and economic

processes that produced the patterning of archaeological remains. However, as Pauketat

(2001) has noted, the processualists were heavily functional in their approach and were

actually interpreting functions or adaptations often quite removed from the material

evidence. Their interpretations regularly constructed abstract “processes” that invoked

ideal systemic laws and causes as befitting a cultural ecology of the archaeological

record. Processualists advanced archaeology with their interpretations of site formation

processes, for example, but practice theory attempts interpretation on the microscale,

closer to the actions that produced the archaeological record. As Pauketat (2001:74)

stated, “the practices are the processes, not just consequences of processes.” This puts

the locus of change not on external and reified systemic processes but rather upon the –– 22 ––

practical actions of individuals and groups themselves. Thus it is important to “locate

the processes of culture change in practices rather than explaining those practices as

consequences of external factors or mechanisms to which people passively and

uniformly respond” (Pauketat and Alt 2005:231). This would include all types of

practices, such as the making of a stone tool, the building of a burial cairn, the holding of

a feast, the exchange of luxury items, or the construction of a fort.

A concept from practice theory that is useful for the understanding of past

human action is the notion of habitus, which Bourdieu (1977:78) described as “history

inscribed into nature.” Rather than mere adaptations to ecological changes, habitus

recognizes how culture and its traditions are embodied and structured within its

practitioners through time. As individuals express the traditions and practices of their

culture, the nature of that habitus leaves its patterns throughout the archaeological

record. And it does so in a historical sequence.

This concept has been applied on the Northwest Coast by other archaeologists

such as Mackie (2003), Grier (2001, 2006), and Mathews (2006). As Mackie (2003:285)

noted in his study of site distributions along the West Coast of Vancouver Island, “To

elaborate upon Bourdieu, the structured dispositions of the habitus lead to a structured

deposition, which itself acts as a structuring deposition” (emphases in original). The

structured dispositions of habitus are the constraining or guiding elements of culture in

the improvisatory and creative acts of past peoples. Rather than treating past

individuals as passive reactors to changes in ecological stimuli, practice theory

incorporates tradition and agency into its operations, reflective of both the passing on of

traditional knowledge and the bricolage it forms as those agents have to improvise in

addressing new conditions. As Grier (2006:104) put it, practice theory offers

“overarching structures [that] provide a spatio-temporal continuity to activity

performance, turning the repeated everyday happening over time into an archaeological

pattern.” That is, practice theory is not just a specific action but is also about how it

connects with or is representative of the broader history or tradition of that practice,

–– 23 ––

which guides that specific action. It is how Bourdieu conceives of history: that which

connects the past to the present, or in archaeological terms, predominantly the moment

of deposition.

Another concept Bourdieu presented was that of the field. Similar to Wolf’s

(1990) use of the term “arena,” as mentioned above, it is the setting in which individuals

strategize and implement tactics in the struggle for resources. The field consists of a set

of social positions structured by, and structuring, their power relationships to one

another. That is, the field is the field of social struggle. Thus, the approach is

materialist; in fact, it is ultimately derived from (Marx 1970 [1845]), when he noted that

“all social life was essentially practical.”

As Marx is known to have turned Hegel upside-down, Bourdieu, it is argued by

Jenkins (1992), flipped the structuralism of Levi-Strauss upside-down, turning his

abstract models onto a materialist base; Bourdieu found structure in material practices

and within each of us as bodily habitus. In this manner, Bourdieu quoted from Marx’s

(1970 [1845]) Theses on Feuerbach for a frontispiece, which stated that “idealism naturally

does not know real concrete activity as such.” Bourdieu’s aims were to provide a

method for analyzing real activities or practices, which explored not structural rules but

improvised strategies of action. In so doing, he not only countered structuralism but

also the solipsistic existentialism of Sartre (1956) by stressing that the nature of habitus,

while subjectively generated originally, is objectively learned through inculcation within

one’s culture and class; thus, both elements of objectivity and subjectivity play a part in

a dialectic of structure and structuration. Both Levi-Strauss and Sartre were then

classed, by Bourdieu as idealistic whereas his approach is materialistic and practical.

Bourdieu also expanded upon Marx’s materialist orientation, by addressing

types of capital beyond economic forms: social capital, cultural capital, and symbolic

capital. Like Wolf’s modes of power, these forms are not simply categories in a

typology––rather they interact dynamically. For instance, Bourdieu had emphasized

that each form of capital is exchangeable for another form of capital. With these

–– 24 ––

concepts, Bourdieu provided insight into other forms of domination beyond the

economic. Accordingly, class struggle is not just against the positions of the upper class,

but also is a struggle against the meanings and significations that are held by the elite,

those symbols and cultural codes that help aid their control through hegemonic

practices.10

A practice approach necessarily integrates power into its analysis; in fact, it can

be combined readily with Wolf’s (1990) modes of power to assess how organizational

power (iii) or structural power (iv) can constrain or control the practices of others in

certain contexts or social fields. Bourdieu emphasized the improvisatory nature of

people furthering their ambitions by enacting practices and options to advance their

own capital or power. This can be seen as the freedom of the actor to either choose

particular practices or even create new ones––although, habitus creates dispositions

toward readily available and historically acceptable practices. On the other hand,

others’ modes of power in the social field are always at play, constraining the available

practices or actions of another.11 The powers of another (ii, iii, or iv) can limit a person’s

actions, but also the power of an individual (i) can enable a greater range or freedom to

pursue certain practices or options, as one can gather capital or organize with others into

higher forms of power (iii & iv), matching or surpassing the organization of the

opposition. Similarly, one’s cultural traditions and dispositions in habitus provide

constraints as to what options are available to pursue. This is the sense in Marx’s (1964

[1852]) statement: “The traditions of the dead generations weigh like a nightmare on the

10. Jenkins (1992:59) noted that Bourdieu was not the first to emphasize practices, as Erving Goffman (1959; also 1961, 1967) had undertaken such studies of institutions from within (especially in his main study on asylums) experiencing their practices, not just studying the documents of the institution, as Foucault (1988) had done.

11. Trigger (1991) titled his major exposition on holistic archaeology “Constraints and freedoms,”finding that processual archaeology had been quite effective in determining external constraints upon behaviour (environmental, technological), while postprocessualists had been better about internal constraints imposed by cultural traditions. Internal constraints are indicated in the practice theory sense of habitus and tradition. However, Trigger (1991:559) also emphasized the improvisatory nature of past peoples by stressing that people are not fully constrained, that there is freedom in their ability to pursue actions to counter both external and internal constraints.

–– 25 ––

minds of the living.” However, Marx also acknowledged that there was a striving to

counter such traditions as well, and it involves pursuing one’s aims as much as possible

within existing constraints:

Men make their own history, but not of their own free will; not under circumstances they themselves have chosen but under the given and inherited circumstances with which they are directly confronted (Marx (1964 [1852]).

Cultural traditions can also be viewed as freeing, if there is a broader range of

options available, although the range of options will inevitably vary through time. In

the past of the Northwest Coast, such abilities to keep the options that one has earned

would have involved constant renegotiation or reproduction of the capital and constant

reenaction of the power one has gained. As Bourdieu (1977:183) discussed for societies

without government, strategies and tactics used to achieve goals are temporary and

must be renewed:

In societies which have no “self-regulating market” (in Karl Polyani’s sense), no educational system, no juridical apparatus, and no State, relations of domination can be set up and maintained only at the cost of strategies which must be endlessly renewed, because the conditions required for a mediated, lasting appropriation of other agents’ labour, services, or homage have not been brought together.

Bourdieu’s point is that the nature of sociopolitical organization affects how

practices are enacted, what practices are available, and the degree of power that can be

achieved––particularly in societies with “no State,” as Bourdieu noted, or anarchic

societies. Practice theory, like Marxism, provides a basis of interpreting from material

evidence readily to broader theorization within social and political structures. Some of

Marx’s theories have been readily taken up by archaeologists (e.g., McGuire 1992;

Gilman 1981; Spriggs 1984b; Kohl 1981; Childe 1951, 1956), because they allowed

interpretations of modes of production and the sociopolitical relations of production.

Because many classes of archaeological remains result from subsistence practices and

economic activity, Marxist theory provides a rubric that ties economy to sociopolitical

structure and ideology. Bourdieu’s practice likewise provides such an overarching

–– 26 ––

theory that allows interpretations of specific practices as patterned by habitus in the

archaeological record to broader interpretations of tradition and the sociopolitical

struggles for various forms of capital. Here, I advocate that anarchism likewise provides

such a rubric that is also materialist or practical in orientation. And, similar to practice

theory, it also does not rely on class struggle for its interpretation, a feature better suited

for capitalist state societies, as Marx had intended.

Anarchism

Anarchism involves a movement and philosophy that has been debated and

worked over for many decades concerning how societies should interact without

overarching forms of government.12 Like Marxism, which similarly had a long history of

theorizing how a communist state would come about, many principles of anarchism

might also be useful for understanding the nature of societies without government, such

as those of the Northwest Coast.

The theory of anarchism, which has been referred to by that label since the time

of Bakunin––a main opponent of Marx in the late 1800s, during the early days of the

International Workers of the World. Others see its anti-government traits in the

American Revolution (e.g., Thomas Paine) or the French Revolution, while some see

these antiauthoritarian principles as extending much further back than that, even

millennia, back to the ancient Greeks or Taoists (Marshall 1993). Anarchist theory

12. Theorists of anarchism include numerous proponents, the first of which are predominantly advanced in the mid to late 19th and early 20th century, such as Pierre-Joseph Proudhon (e.g., 1890, 1972, 1979; Noland 1967, Woodcock 1962); Mikhail Bakunin (e.g., 1916, 1950; Morris 1993; Maximoff 1964; Carr 1937); Peter Kropotkin (e.g., 1902, 1910, 1946, 1996a [1901], 1996b [1910 -1915]; Morris 2004); Elisée Reclus (e.g., 1886; Clark and Martin 2004); and Emma Goldman (1917). In many quarters, other major thinkers are regarded within the scope of anarchism, such as William Godwin (1976 [1796]), Max Stirner (1907), and Leo Tolstoy (1990). Into the later 20th century, the theory of anarchism further developed through Rudolph Rocker (1998), Colin Ward (1973), Noam Chomsky (2005), and Murray Bookchin (1971, 1991), among others. Recently, theorists have also adapted anarchism in light of its affinity or relevance to postmodern and poststructuralist thinkers such as Foucault, Derrida, and Lacan (May 1994, 2001; Newman 2001; Call 2003). In anthropology, anarchism has been discussed by Perry (1978), Clastres (1987, 1994), Barclay (1982, 1997, 2003), Graeber (2002, 2004, 2007), and Morris (2005).

–– 27 ––

emphasizes principles of network organization (as opposed to centralized hierarchies),

gift economies, local autonomy, and actively opposes the rise of centralized authorities.

Anarchism also provides a theory of history that is an alternative to, if not contrary to, a

Marxist one. In the following, I discuss the core principles of anarchist theory before an

assessing an anarchist theory of history.

Core Principles of Anarchism

Due to long history as well as its acephalous nature, anarchism actually

comprises a broad corpus of ideas with a variety of strains. And, unlike Marxism, which

is tightly associated with primarily one individual (especially in name), no one thinker is

predominant. Just as its theoretical outlook suggests, much of anarchist thought and

practice encourages variation and is oriented to local circumstances. There are

individualist (“egoist”) strains, collectivist, anarcho-syndicalist, neo-primitivist,

ecological anarchist, and more; David Graeber (2004) has noted that these anarchist

strains are named not after philosophers, but after practices or principles.13 True to

anarchist beliefs, no one thinker is dominant. It has been said that one does not even

have to know who Kropotkin, Bakunin, Rocker or Bookchin are to be an anarchist––a

similar statement could not be said for Marxism. Rather than canonical texts, there is

13. “Now consider the different schools of anarchism. There are Anarcho-Syndicalists, Anarcho-Communists, Insurrectionists, Cooperativists, Individualists, Platformists.... None are named after some Great Thinker; instead, they are invariably named either after some kind of practice, or most often, organizational principle. (Significantly, those Marxist tendencies which are not named after individuals, like Autonomism or Council Communism, are also the ones closest to anarchism.) Anarchists like to distinguish themselves by what they do, and how they organize themselves to go about doing it. And indeed this has always been what anarchists have spent most of their time thinking and arguing about. Anarchists have never been much interested in the kinds of broad strategic or philosophical questions that have historically preoccupied Marxists—questions like: Are the peasants a potentially revolutionary class? (Anarchists consider this something for the peasants to decide.) What is the nature of the commodity form? Rather, they tend to argue with each other about what is the truly democratic way to go about a meeting, at what point organization stops being empowering and starts squelching individual freedom. Or, alternately, about the ethics of opposing power: What is direct action? Is it necessary (or right) to publicly condemn someone who assassinates a head of state? Or can assassination, especially if it prevents something terrible, like a war, be a moral act? When is it okay to break a window? To sum up then: 1. Marxism has tended to be a theoretical or analytical discourse about revolutionary strategy. 2. Anarchism has tended to be an ethical discourse about revolutionary practice” (Graeber 2004:5-6).

–– 28 ––

instead an adherence to a set of principles that guides much of anarchism and its

practices and these provide connections among the various strains. These principles

include: individual and local autonomy, voluntary association, mutual aid, network

organization, communal decision-making, direct action, justified authority, and

decentralized forms of social organization (in addition to resistance to centralized forms,

such as that of states). Here I focus on the principles that are most relevant for an

archaeological application.

Individual and local autonomy

Within anarchic societies, the locus of control is not within any center but rather

is distributed more broadly throughout the society. The centers of control are stronger

at the scale of the individual, family, and household. According to anarchist theorists,

society should be organized from the bottom-up, with groups voluntarily associating

with other groups in broader confederations at larger scales. The locus of control

remains on the local scale. Proudhon (1890) had argued for “the autonomy of the

private reason, originating in the difference in talents and capacities.” While anarchists

advocate for autonomy, it does not mean atomism, which implies independent agents

concerned for their own affairs. Rather, autonomy conveys personal and local group

freedom but with extensions of cooperation through voluntary association.

Voluntary association

Anarchism is closely associated with the furtherance and enhancement of

individual and local freedom and expression. An emphasis on autonomy fits with a

principle of voluntary association. Instead of a state determining the relationships of its

constituents, anarchists prefer individuals to voluntary associate with other groups for

tasks or shared interests. Even within modern state societies, Kropotkin (1927:66; cited

in Morris 2004:69) often pointed to the voluntary societies that are “constituted everyday

for the satisfaction of some infinitely varied needs of civilized man,” including trade

–– 29 ––

unions, professional and scientific societies, or the Red Cross. These societies were

formed without decree from a centralized government for their formation or for

individuals to necessarily participate; rather, these form from shared social needs and

interests. With smaller societal scales, voluntary association is a principle applied when

individuals or groups opt to form an alliance or to participate with others for collective

endeavors that local groups could not accomplish on their own. Voluntary associations

will tend to ensure that tasks or activities conducted meet the needs of those involved. If

the interest or need in such an activity wanes or becomes unnecessary, the association

will dissipate. Voluntary association with a group or activity also will tend to ensure

positive relationships between those involved. If relations turn negative, the association

or relationship can simply be severed. Therefore, to maintain relationships and such

associations, a principle of mutual aid connects the shared interests of those involved.

Mutual aid

In a society developed on these lines, the voluntary associations which already now begin to cover all the fields of human activity would take a still greater extension so as to substitute themselves for the state in all its functions. They would represent an interwoven network, composed of an infinite variety of groups and federations of all sizes and degrees, local, regional, national and international––temporary or more or less permanent––for all possible purposes: production, consumption and exchange, communications, sanitary arrangements, education, mutual protection, defence of the territory, and so on.––Peter Kropotkin (1910)

Mutual aid is a driving principle for connecting individuals and groups in

cooperative endeavors. It contributes to a self-organizing, bottom-up form of social

interaction. Self-organization refers to the ability of groups to organize organically,

without a centralized authority governing the organization of groups. Anarchists

believe that no central authority is necessary to accomplish any given endeavor. To

Kropotkin, the cultural evolution of humankind attests to this, as humans have survived

and proliferated for most of their evolutionary history in societies without government.

While, as humans, we may not have always been formally governed, we have always

–– 30 ––

been social. This was explicitly a focus of Kropotkin in Mutual Aid (1902). In the

decades after Darwin (1970 [1859]) published The Origin of Species there was a

proliferation of ideas, influenced by Thomas Malthus’ (1798) economic principles which

emphasized competition. Thomas Huxley (1888) called it the “struggle for existence,”

while Herbert Spencer (1891) extolled the “survival of the fittest.” Kropotkin viewed

such notions as attempts to buttress capitalism, and, while recognizing that struggle and

conflict were important factors in evolution, wanted to emphasize that cooperation was

also a critical factor.

Stephen Jay Gould (1988:11) stated that “Kropotkin’s basic argument is correct,”

noting that he perhaps overemphasized cooperation, but primarily to balance the Social

Darwinist tendency that overemphasized competition. Gould (1988:16-17) remarked

how Darwin had investigated the tropics, while Kropotkin, soon after reading The Origin

of Species, conducted seasons of geographical explorations in Siberia. Their experiences

led each to separate conclusions. Whereas Darwin saw the fight over resources in the

plentiful, tropical environs, Kropotkin witnessed organisms whose primarily struggle

was against the environment. From this experience, Kropotkin postulated that there

were two types of struggles: one that engaged each organism against another, or

competition; the other being individual organisms in coordination against the

environment:

Two aspects of animal life impressed me most during the journeys which I made in my youth in Eastern Siberia and Northern Manchuria. One of them was the extreme severity of the struggle for existence which most species of animals have to carry on against an inclement Nature; the enormous destruction of life which periodically results from natural agencies; and the consequent paucity of life over the vast territory which fell under my observation. And the other was, that even in those few spots where animal life teemed in abundance, I failed to find––although I was eagerly looking for it––that bitter struggle for the means of existence, among animals belonging to the same species, which was considered by most Darwinists (though not always by Darwin himself) as the dominant characteristic of struggle for life, and the main factor of evolution (Kropotkin 1902).

This aspect of evolution is repeatedly returned to, especially among

anthropologists. For instance, Quigley (1971), Read (2003), and Isaac (1978) have argued

–– 31 ––

for the importance of cooperation in hominid evolution. In an ethnography of the Kung

San, Alan Barnard (1993) attempted to use Kropotkin’s concept of mutual aid as a better

depiction of their life-ways, particularly as “an alternative, very much non-Marxist view

of primitive communism.” Richard Lee’s (1988) description of San foragers as

exemplifying primitive communism or a communal mode of production provided an

interesting perspective on the economics of communalism, however, Barnard argued

that a foraging ethos extended well beyond the economy. Asserting there was in fact a

“foraging mode of thought,” he described communalism as a mode of interaction that

determines social relations and is ideologically embedded (Barnard 1993). These social

interactions of mutual aid and cooperation form, through repeated engagements with

others, networks of organization.

Network organization

Network forms of organization do not exhibit centralized or hierarchical forms.

Instead of channelling information downward from those in the upper echelons of a

pyramidal structure, networks exhibit rich horizontal linkages. This does not mean that

all nodes in a network are equal, as placement within a network can engender certain

advantages. However, in networks, the flows of information are more open and

organizational responses generally occur at a local level. Among many groups,

including the Coast Salish, many social networks are created along lines of kinship. As

Read (2003) argued, kinship––despite its namesake––is not really about literal kin, nor

genealogical ties, but about how cultural groups ascertain who is related. Kinship is a

method of extending one’s familial relations beyond those immediately related

biologically. Read (2003) viewed kinship as a mode of ready self-organization that

promotes cooperative behavior, a point that parallels Mithen’s (1996) argument that

emphasizes sociality throughout human evolution, more so than technical or

environmental knowledge.

Mutual aid or cooperative endeavors are seen by anarchists as the core dynamic

for the self-organization of groups and for the linking of those local corporate groups –– 32 ––

into larger community and regional networks. Anarchic organization is not driven by

singular leaders, but rather are generated and structured by the needs of the people

involved. According to Bookchin (1991), the practical needs of individuals within local

groups are the medium of organization, and organizations can respond as immediately

as the need arises, which he described as “an ordering and structuring force.” Colin

Ward (1973:11) stated that, in part because of this emphasis on self-organization, this

revealed that anarchism is not utopian (as many have classed it): “far from being a

speculative vision of a future society, it is a description of a mode of human

organization, rooted in the experience of everyday life.” He referred to anarchism as a

“theory of spontaneous order” (Ward 1973:28). Likewise, Bakunin (1950 [1872]:18)

himself, noted that “liberty must establish itself in the world by the spontaneous

organisation of labour.”

Network forms of organization, as defined by Podolny and Page (1998:59), are

“any collection of actors (N>2) that pursue repeated, enduring exchange relations with

one another and, at the same time, lack a legitimate organizational authority to arbitrate

and resolve disputes that may arise during the exchange.” Networks are in opposition

to market-based or hierarchical relations. Market relations are short lived, existing only

for the period of exchange––the end of the exchange effectively ends the relationship,

whereas networks maintain those relationships. Hierarchical relations exhibit a “clearly

recognized, legitimate authority [that] exists to resolve disputes,” whereas networks

exhibit conditional and situational authorities (Podolny and Page 1998:59). Moreover,

they noted that network forms of organization adapt more quickly to changes due to

faster lines of communication than those found in centralized forms. Not only does

information travel faster, but it also conveys “richer, more complex information” that

also is subject to a wider array of offered responses from various nodes in the network,

as opposed to the narrow options to be delivered from the managers in centralized

forms of organization (Podolny and Page 1998:62-63).

Archaeologists have also analyzed principles of self-organization as important

–– 33 ––

for understanding the past. For instance, David Braun and Steven Plog (1982) argued,

using studies of the U.S. Southwest and Southeast, that “tribal social networks” were

adaptive organizations that could respond to environmental resource instabilities. For

Bronze Age Europe, Gilman (1981) stated that labour-intensive projects (such as plow

agriculture or irrigation canals) would not necessarily have been led by centralized

elites, as each of the examples, considering the scale, could have been locally organized

and maintained. Through networks, major problems and projects could be effectively

addressed in decentralized fashion.

Decentralization and antiauthoritarianism

True progress lies in the direction of decentralization, both territorial and functional, in the development of the spirit of local and personal initiative, and of free federation from the simple to the compound, in lieu of the present hierarchy from the centre to the periphery.––Peter Kropotkin (1910)

If mutual aid is something anarchists support, authority is something they

oppose. Sebastien Faure wrote that “Whoever denies authority and fights against it is an

anarchist” (Woodcock 1962:9). Saul Newman (2001:37) observed that “History, for

anarchists, is this struggle between humanity and power”––such is anarchism’s focus on

antiauthoritarianism, and, in particular, its rejection of state authority. According to

Foucault (1980, 1997), all social relations of dominance and coercion embody relations of

power. Moreover, power refers not to an abstract entity or essence, but rather refers

only to the nature of relationships. He emphasized that power was never total. If power

were totalizing, it could no longer be considered as power. Power relations indicate that

some degree of freedom is able to be deployed by those actors.14 One could say that

14. “[P]ower relations are thus mobile, reversible, and unstable. It should also be noted that power relations are possible only insofar as the subjects are free. If one of them were completely at the other’s disposal and became his thing, there wouldn’t be any relations of power. Thus, in order for power relations to come into play, there must be at least a certain degree of freedom on both sides.... But the claim that ‘you see power everywhere, thus there is no freedom’ seems to me absolutely inadequate. The idea that power is a system of domination that controls everything and leaves no room for freedom cannot be attributed to me” (Foucault 1997:291-93).

–– 34 ––

Foucault presented a social science variant on Newton’s third law, “To every action

there is an equal and opposite reaction,” which would be: To every application of

authoritarian power there is an opposing resistance.

Much of the anarchist literature on power concerns the state, although in this

study, I am particularly concerned with the application of power in non-state societies.

A few anarchist theorists extended their notions of power to non-state cases as well.

Proudhon, for instance, noted that “All parties without exception, in so far as they seek

for power, are varieties of absolutism” (Woodcock 1962:18). Thus, through an anarchist

perspective, authoritarian power is something to be challenged. Newman (2001:37)

praised the anarchist critique of Marxism for opening the door to wider examinations of

noneconomic forms of power. Proudhon and Newman, however, both exhibit a rather

shallow conception of power, limiting it to a vertical notion of power, expressed from

top to bottom. Anarchists recognize power in solidarity, or what could be called

horizontal power, the power of solidarity. Opposition to authority is often described as

the lifeblood of anarchist revolutionaries and, accordingly, helps to sustain anarchic

communities. As Bakunin (1970 [1916]:35) said, concerning resistance to authority, “This

is the sense in which we are really Anarchists.” However, outspoken Bakunin was

about authority, he did not reject it entirely. Rather, it is more accurate to state that

anarchists maintain an opposition to authoritarianism.

Justified authorities

Does it follow that I reject all authority? Far from me such a thought. In the matter of boots, I refer to the authority of the bootmaker; concerning houses, canals, or railroads, I consult that of the architect or engineer. For such or such special knowledge I apply to such or such a savant. But I allow neither the bootmaker nor the architect nor the savant to impose his authority upon me. I listen to them freely and with all the respect merited by their intelligence, their character, their knowledge, reserving always my incontestable right of criticism censure. I do not content myself with consulting authority in any special branch; I consult several; I compare their opinions, and choose that which seems to me the soundest. But I recognize no infallible authority, even in special questions; consequently, whatever respect I may have for the honesty and the sincerity of such or such an individual, I have no absolute faith in any person. Such a faith would be fatal to my reason, to my liberty,

–– 35 ––

and even to the success of my undertakings; it would immediately transform me into a stupid slave, an instrument of the will and interests of others. ––Bakunin (1970 [1871]:32)

Anarchists recognize “authorities” about a matter for their knowledge or

experience. Bakunin (1970 [1871]:32) stated “I bow before the authority of special men

because it is imposed upon me by my own reason.... Therefore there is no fixed and

constant authority, but a continual exchange of mutual, temporary, and, above all,

voluntary authority and subordination.”15 This view of authority as being something

rooted in specialized knowledge or skills has commonly been noted anthropologically

among many cultures. Among the Coast Salish Puyallup-Nisqually, Marian Smith

(1940) noted that warriors, or war chiefs, were given specific powers over villages,

related to war activities, but only for the duration of hostilities. The designation of such

authority must be carefully and situationally justified, lest it become authoritarian.

Noam Chomsky summarized this anti-authoritarian stance as a core expression of the

anarchist principles:

Anarchism, in my view, is an expression of the idea that the burden of proof is always on those who argue that authority and domination are necessary. They have to demonstrate, with powerful argument, that that conclusion is correct. If they cannot, then the institutions they defend should be considered illegitimate. How one should react to illegitimate authority depends on circumstances and conditions: there are no formulas (Chomsky 1996).

An Anarchist Theory of History

The conception of history was one of Bakunin’s main points of contention with

Marx––while their end-goals for their philosophies might appear similar (aiming for

15. Bakunin (1984:239) also referred to this distinction as between natural and artificial authority. The former is justified as an expression of natural human relationships, whereas artificial authorities are imposed through institutional structures. Newman (2005:172; see also 2001:38-41) considered this distinction a “major theoretical achievement of anarchism.” No longer could one claim “what replaces the state?” as the anarchists’ conception of power is not tied to the state or the “social contract,” but concerns human relationships (Newman 2001:40).

–– 36 ––

communism)––their approaches to achieve that solution were widely different, and

according to Bakunin, would lead to wildly different outcomes. Whereas Marx

advocated a communist state to rule all in an egalitarian fashion, anarchists opposed any

imposition of the state. Those like Bakunin criticized the notion that a “Dictatorship of

the Proletariat” would rule over others in egalitarian fashion until a “withering away of

the state” could occur––they thought it to be simply naive in its conception of power.

Once leaders acquire power, Bakunin and other anarchists maintained, they will

struggle to hold onto that power, whether or not they were originally members of the

proletariat.16 Stalin later provided a ready example of Bakunin’s critique of such a drive

to maintain power as the head of a communist state. Like Tolkien's “ring of power,”

acquiring sociopolitical power easily leads to corruption and creates a need to do all that

one can to maintain that power, even when one might corrupt into something like

Golum. This anarchist emphasis on incorporating power into analyses of social systems

predates Foucault by well over a century. In fact, some have claimed that Foucault was

(or could be called) a postmodern anarchist (e.g., May 1994, Newman 2001, Call 2003),

although it is debatable whether he himself made such a claim. Regardless, his work

has been useful to anarchist philosophy in many respects, particularly for his microscale

explication of the fingers of power through institutions and fields from governmentality,

prisons, medicine, science, to individual sexuality. If there is a criticism from the

anarchist perspective, it is that the type of power Foucault described is largely restricted

to institutional, bureaucratic forms, where knowledge equals power. Instead, as

Graeber (2004:71-72) pointed out, the real “brute force” power of the state is always close

16. This strident anarchist message, nonetheless, was picked up in later critiques of Marx. Gramsci (1971 [1929-1935]), for instance, while Marxist, added the concept of hegemony to Marxist interpretation, which showed that not only did the ideology, or the “opiate” (as Marx referred to religion) need to be cracked, but that revolution required opposing the material implementation of power that resides throughout culture in hegemonic practices (this is termed by Kurtz [1996, 2001] in an anthropological adaptation as “hegemonic culturation”). To paraphrase simply, ideology is akin to mythology, while hegemony is similar to ritual practice. Ideology can be countered and needs to be if one is to resist it, but the hegemonic practices must also be faced with counter-hegemonic actions.

–– 37 ––

at hand and ready to reduce any threats to the status quo.17

These differences in understanding power led Alan Carter (1988, 1989, 2000) to

detail two varying theories of history. Marx, with his emphasis on economy, described

power as arising bottom-up from basal economic forces, whereas Bakunin and other

anarchists saw authoritarian power as originating at the top and working its way down

through chains of command through sociopolitical forces. They argued that

authoritarian power is ultimately centralized and acted upon at sociopolitical apexes,

even if that power was acquired through the control of economic capital. Long before

later theorists would try to update or reformulate Marxism (e.g., Gramsci 1971

[1929-1935]; Althusser 1969, 1986) to account for its overly economic or vulgar

applications, anarchists had already made these criticisms. The problem was not the

focus on economic capital—indeed, many anarchists respected Marx’s powerful

exposure of capitalist dynamics. Rather, they simply criticized its weak conception of

power. The important difference in an anarchist perspective for Kropotkin (1927:150)

was that “it attacks not only capital, but also the main sources of the power of

capitalism: law, authority, and the state.”

There is another significant difference between Marxist and anarchist theories of

history: Marxism is teleological, while anarchism is not. According to early Marxists, an

ideal communist state would eventually arise, after the stage of capitalism. Societies

existing prior to this stage were labeled “pre-capitalist” (Marx 1965 [1857-1858]). For

17. "Academics love Michel Foucault’s argument that identifies knowledge and power, and insists that brute force is no longer a major factor in social control. They love it because it flatters them: the perfect formula for people who like to think of themselves as political radicals even though all they do is write essays likely to be read by a few dozen other people in an institutional environment. Of course, if any of these academics were to walk into their university library to consult some volume of Foucault without having remembered to bring a valid ID, and decided to enter the stacks anyway, they would soon discover that brute force is really not so far away as they like to imagine––a man with a big stick, trained in exactly how hard to hit people with it, would rapidly appear to eject them. In fact the threat of that man with the stick permeates our world at every moment; most of us have given up even thinking of crossing the innumerable lines and barriers he creates, just so we don’t have to remind ourselves of his existence. If you see a hungry woman standing several yards away from a huge pile of food––a daily occurrence for most of us who live in cities––there is a reason you can’t just take some and give it to her. A man with a big stick will come and very likely hit you. Anarchists, in contrast, have always delighted in reminding us of him” (Graeber 2004:71-72).

–– 38 ––

anarchists, the term “pre-capitalist” was teleological because it presumes capitalism was

inevitable––a position they did not accept, just as they did not accept the communist

state as a future utopia. Rather, anarchism was anti-Progress, in the Victorian sense, and

it was closer to Darwin’s non-progressive view of evolution than was Marxism.

Integrating Anarchism, Practice, and Power

These three theoretical orientations––anarchism, practice, and power––may

appear disparate, but these conjoin in a manner that aids the utility of each. The theory

of anarchism provides principles of social organization among small-scale, largely

autonomous societies without overarching governments. It is these principles, applied

to the needs and desires at hand, that organize their economy, sociality, and ritual.

These principles, applied in varying local expressions, reflect how anarchic groups

organized their daily practices. Practice theory provides a way to understand how these

practices are embodied, through history and habitus, into long-lasting, structuring

traditions. At the same time, practice theory affords understandings of individual

agency, of how practices are strategically and tactically selected and implemented in an

improvisational manner that advances an individual’s or a group’s pursuit of capital

within a social field. Amassed capital contributes to a concentration of power. Finally,

Wolf’s (1990) modes of power provide a way to investigate the dynamic dimensions of

power from the individual to the level of interacting societies.

Each of these approaches within our framework of power, practice, and

anarchism are deserving of more detail, and in the chapters that follow, I examine them

with respect of various aspects of Coast Salish warfare as we proceed through the

archaeological, ethnohistorical, oral historical, and ethnographic evidence.

–– 39 ––

Chapter III: The Anthropology and Archaeology of Warfare

Warfare in non-state societies is often portrayed in either Hobbesian or

Rousseauian terms: either warfare was a constant presence, due to the lack of a

Leviathan to maintain order; or warfare was limited to ritualistic ceremonies of no real

material consequence, if it occurred at all. Warfare, according to the latter view, is the

bane of states. Anthropologically, theories of the origin of warfare can be placed mostly

in one or the other camp, but, as McGuire (2001) pointed out, such either/or theories are

of little explanatory use. For one, those are too generalizing as to have limited

applicability to local and historical circumstances that condition and shape the decisions

to engage in warfare. Instead, theories are needed incorporate local cultural dynamics

but also historical and environmental conditions. In this chapter, I consider major

anthropological and archaeological treatments of warfare and situate the theoretical

framework developed here for an inquiry into Coast Salish warfare––namely power,

practice, and anarchism––to show the advantages of such a multitiered and scalar

approach. Anthropological theories of warfare centre around the origins of warfare and

its causes, so the first two sections discuss those. In the third section, I incorporate the

larger discussion of warfare in the broader social science of political philosophy,

relevant to this discussion, specifically Marxist and anarchist approaches to warfare.

The Concern with Origins

Otterbein (2004) maintained two separate origins for warfare, involving “two

types of military organization”: The first, two million years ago “at the dawn of

humankind,” and the second among agricultural peoples that achieved early statehood.

The point, here, is that Otterbein stressed “organization,” which matches nicely the

development of power (iii) as outlined by Wolf (1990). Individual power (i) is and has

–– 40 ––

always been present in any individual, although relative and constantly in flux or in

need of maintenance. Interpersonal or relational power (ii) is the deployment of one’s

power towards another; consequently, it also has always been present among humans,

however, physical enactions towards another are not regarded as warfare, but as

violence. It is Wolf’s (1990) third mode, organizational power (iii), that is the mark of

warfare. It is the recognition that warfare is a cultural practice enacted by groups.

Wolf’s model is beneficial as it connects warfare with lesser scales of power that reside

in and between individuals, however, it also shows how power increases not only in

scale, but in dimension. Furthermore, Wolf’s framework also adds another dimension,

which is structural power (iv), the point at which organizational power (iii) controls or

can alter, even demolish, the settings of interaction. Many anthropological definitions of

warfare contain this notion of organization (iii) as a criterion for warfare. The definition

of what constitutes war also has a long history.

Harry Turney-High (1949), an early analyst of warfare, maintained a distinction

called the “military horizon” which separated “primitive war” from “true war.”

Whereas the former required the recruitment of volunteers, the latter had command

structures and campaigns; Lawrence Keeley (1996) later critiqued Turney-High for his

treatment of warfare and states, which elevated modern warfare and seemed to

insinuate that primitive warfare was a sport. In general, definitions of warfare are

subject to interpretation. Even today, offensive attacks are often said by the initiators to

be defensive actions, or acts of “preemption” against future attacks, carried out under

the auspices of a “Department of Defense.” Analytically, some basic terms can be

established despite multifarious interpretations of attackers and defenders. Despite his

elevation of military warfare, Turney-High (1971 [1949]:5), in Primitive War, recognized

matter-of-factly: “War is war.” Perhaps it does not need further definition:

We are not here speaking of anything so complicated as an explanation of polyandry among separated peoples, nor the similarities or dissimilarities in the ceramic complex of the Americas and New Stone Age Europe. The art of war or the artlessness of fighting are so simple throughout time and over the face of the world that the discussion could

–– 41 ––

be made very monotonous (Turney-High 1971 [1949]:24).

His work being a “primer” on the subject, he did provide a definition of warfare

as “a social institution fulfilling a variety of motives, ending in many ways, evoking

many emotions. The central fact of military theory is that war is a sociologic device, and

weapons are merely tools used to facilitate its practice” (Turney-High 1971 [1949]:5).

This has some congruence with anthropological treatments of warfare, as Malinowski’s

(1936:444) definition indicated, referenced earlier (see page 7), regarding it as “use of

organized force between two politically independent units.”

From perhaps the first anthropological conference on the anthropology of

warfare (Fried, Harris, and Murphy 1968), Margaret Mead (1968:215) offered another

definition: “Warfare exists if the conflict is organized, and socially sanctioned and the

killing is not regarded as murder.” Mead forefronted the organizational aspect of

warfare but also recognized its distinction from “murder,” which would only be

interpersonal power (ii); although it should be noted that interpersonal squabbles and

murder are often the fires that escalate into full-scale warfare (iii), as it has in the Coast

Salish past.

In discussing the results of a later conference on the anthropology of war (Haas

1990), McCauley (1990:1) defined war as a “a subset of human aggression involving the

use of organized force between politically independent groups.” Mead’s (1968:215)

earlier definition is perhaps more encompassing, highlighting the agential or active

nature of warfare as well as emphasizing that such acts of war are socially decreed,

although McCauley’s treatment stressed the autonomy of the groups engaged in a

conflict, just as Malinowski’s (1936:444) definition. As regarded by analysts of

international politics (e.g., Wendt 1992), any scenario of warfare, whether between

polities or within a polity (as in civil wars or revolutions) consists of autonomous

groups. The presence of conflict itself minimally indicates assertions of autonomy, even

as one or both (or more) attempt to dominate the other group(s). To describe

international conflicts, they use the term “anarchy,” in that no overarching institution

–– 42 ––

has control or authority as warfare, by its very enaction, indicates an autonomous action,

whether it be the rejection of another’s claim (external to a polity) or the rejection of

one’s authority (internal to a polity). The success or failure of the acts of warfare

determine the relations of authority, if any, afterwards. For our purposes, warfare

consists of violent and coercive practices conducted in organized means (iii) against

other autonomous groups.

Scalar Approach to Power

There is more to Otterbein’s (2004) argument for the origins of warfare. These

moments of militaristic organization occurred in response to specific conditions. He

maintained that, two million years ago, warfare proliferated among big-game hunters.

Hunter-gatherers, in general, engage in warfare less often than agriculturalists,

horticulturalists, or herders, even if they exhibited many episodes of interpersonal

violence (ii) (Keeley 1996:186, Table 2.2; Otterbein 1999, 2004:81). Since early hominid

days, hunter-gatherers exhibited traits of cooperation, Otterbein (2004:39) maintained. It

is what enabled humans to be successful:

Early humans were cooperators. Among early hominids (austrolopithecines and early members of the genus Homo) cooperation was the key to survival. It permitted them to attack other animals and as well as to repel attacks by them.

This aspect of human evolution has been noted often. For instance, Quigley

(1971) and Isaac (1978, 1983) have argued for the importance of cooperation in hominid

evolution; Kurland and Beckerman (1985) argued that cooperation was more important

than labour and tool use, particularly in the savannah environments where cooperative

sharing of information about highly distributed resources would have been critical.

Read and LeBlanc (2003) argued that cooperative behaviour––or what we could term,

after Kropotkin (1902), mutual aid––is a defining trait that allowed for the emergence of

Homo. In contrast, Old World monkeys and many African ape species (chimpanzees

excluded) exhibited more individualistic behavior that required a reversion to smaller

–– 43 ––

groups, according to Otterbein (2004:39). However, this cooperative behavior arose in

conjunction with a new conceptual system of kinship, which allowed a determination of

who among others is likely to be cooperative. Kinship, then, overtakes genealogical ties,

in advancing reproductive fitness, as cooperation is pursued with those that are

determined to be related culturally, whether biologically related or not. Kinship, in

Otterbein’s (2004:39) argument, is a mode of ready self-organization that advances

cooperative behavior, a notion consistent with Mithen’s (1996) argument that sociality is

the most advanced component of early hominid minds, more so than technological or

environmental knowledge.

Cooperation is important, as Otterbein (2004:39) noted, that “defense and attack

depend upon cooperation.” This hints at a complex conception of interaction intrinsic to

warfare, that cooperation and conflict act at once, depending on which scale of analysis

is employed: “Conflict and cooperation are the opposite sides of the same coin. When

there is conflict between groups, there is cooperation within the groups in conflict”

(Otterbein 2004:46). This is a notion first advanced by the sociologist Lewis Coser

(1956:87), who proposed that “conflict with out-groups increases internal cohesion.”

Furthermore, cooperation is important in that “Group cooperation led to

fraternal interest groups, the first military organization” (Otterbein 2004:39; also

Otterbein 1989). Big-game hunters differed from other hunter-gatherers in their

formation of fraternal interest groups, where leaders gather followers and form factions

of kinsmen. His conception is opposed to egalitarian hunter-gatherers that emphasize

smaller game and gathering more prominently. Big game hunting, with its focus on a

collective hunting practice, produced a culture that was oriented towards killing large

prey, which required cooperation and tightened the bonds of those involved in the

hunt. Given their bonds, fraternal interest groups became the “key elements in

situations of conflict” (Otterbein 2004:45):

If one individual challenges another, and if kinsmen of the challenged person are in the vicinity, they will come to his aid. If the challenger’s kin are also in the vicinity, they in turn will come to his aid. Conflict has escalated, and now it is not two individuals but two fraternal interest

–– 44 ––

groups confronting each other (Otterbein 2004:45).

Here, Otterbein showed how violence between individuals (ii) escalates into

organized warfare between individuals (iii). His description of fraternal interest groups

relates to another major theory about the origin of warfare by Raymond Kelly (2000).

He argued that warfare occurs because of “social substitutability,” where an attack on an

individual (ii) is perceived as an attack on one’s group (iii); therefore, warfare occurs

when societies become segmented, as with fraternal interest groups.18 Kelly’s (2000)

“segmentary societies” is perhaps the more accurate term to use. While “fraternal

interest groups” do encapsulate a majority of culture areas, the term would not be

accurate for others, such as the Coast Salish, where allies are formed not just with

kinsmen but affines that often were not close but distantly located. In any case, both of

these theorists highlight the role of organized factions, if they differ on when those

formations occur; Otterbein (2004) argued for early hominids and even chimpanzees

having fraternal interest groups, while Kelly (2000) argued that this began 20,000 years

ago in the Sudan, with archaeological evidence prominent after 10,000 BP, similar to

Haas (2003).

In total, Otterbein (2004) viewed a triad of factors involved in the formation of

war: fraternal interest groups, weapons, and a focus on hunting. These could be

categorized as factional social relations, means, and cultural practice. The latter, he

argued, is a cultural practice oriented towards big-game hunting, already a military-like

organization focused on game. As Luckert (1990) concluded, it is simply one more step

to go after even greater prey than big game: other humans. For those concerned with

status, particularly those in segmented groups where leaders compete for followers,

attacking the most formidable prey––other humans––is an avenue for such

opportunities. Luckert (1990), and others (Kroeber and Fontana 1986; Ehrenreich 1997),

18. Flannery and Marcus (2003) have put Kelly’s (2000) framework into archaeological practice by determining the segmentation of groups in early Oaxacan states, leading to the formation of Zapotec state.

–– 45 ––

have argued that in the decline of big-game or hunting as an economic emphasis,

individuals turned to the “hunting” of humans to gain status.

There has been an overall scheme related to the prey/hunter dichotomy.

Barbara Ehrenreich (1997) argued that the passions for warfare developed in early

hominids in response to the threats of great predators.19 Fighting back against such

predators instilled great group unity and cooperation and directed, even sacralized, the

act of killing the beast, the other, whether animal or human prey. That is, the ideology

develops in fear of attack of injury upon them––the threat of death. This dialectic of life

and death has a significant imprint for early religions, and for this reason it often serves

as a path for rites of passage, where initiates symbolically are put to death to be reborn

into a new social status.20

Maurice Bloch (1992) intended to search for “archetypes” of such religious

rituals, which he considered to be universal constructs that served to play a part

between human mortality and the seeming permanence of the group or societal

structures. He viewed rituals as a dialectic between life and death, mortality and

eternality; in other ways, a dialectic between eating and being eaten, the dynamics of the

food chain.21

19. Ehrenreich (1997) developed her argument concerning the anthropological history of warfareultimately as a social critique to argue that such segmentary divisions and ritualistic feelings (its powerful ideology and “contagious” fervor and nationalism) and preparations for war continue to influence the actions of politicians in international relations. While many treatments of war turn to concerns with economy, she emphasized its nearly religious appeal in nationalism, which she connects to these early human concerns in bonding through fear or fervor against other groups.

20. Rituals have long been analyzed as having a dialectic structure: van Gennep’s (1960 [1908]) separation, transformation, and reaggregation into society; Turner’s (1969) elaboration and emphasis on the middle stage of liminality where people engage the sacred and participate in antistructural communitas. Following their basic structure, Bloch (1992) added the importance of violence in such rituals, which those theorists had underplayed. Invocations of death and violence present an alternate logic: for mortal life, birth is the event which precipitates life and growth, whereas in ritual life it is death (as in sacrifice) that burgeons immortality or transcendence. Due to this logic, the symbolic participations in death and violence are the result of actors aiming for transcendence in both religion and politics, and so it is not from an innate propensity towards violence as some have argued (e.g., Girard 1986; Burkert 1996).

21. Theorists have also argued that ritual sacrifice has its origins during an early period in which humans were prey to greater-than-human beasts. Luckert (1991) argued that the origin of ritual sacrifice involved the act of a hunter or scavenger, who satisfied predators (or “gods” since they were the predators, greater-than-human realities) by offering a share of their

–– 46 ––

Among the Jivaro, Elsa Redmond (1994) had noted a similar rite of passage in

which children do not become adults, but rather become warriors. In their ritual, the

children attack a sloth, treating it symbolically as a human enemy. The hunting practice

is used as a substitution for the practice of warfare. Adulthood is no longer just

becoming a predator, but a predator of human enemies. In some cases, the emphasis on

warriors is connected with cannibalism, as among the Maori (Vayda 1960a, 1960b) and

among the Jivaro (Redmond 1994:49).

Bloch (1992) examined a ritual from the Orokaiva in New Guinea that he

considered crucial for understanding the importance of the shift from prey to hunter.

During the ritual, feathered- and pig-masked adults chase after boys and girls yelling

“Bite! Bite! Bite!” They surround and herd the children onto a platform normally used

for slaughtering pigs. Then, all the children are draped and hidden, and taken to an

isolated hut outside the village; symbolically, the children have been eaten by the spirits.

After a long period of ordeals and training, the children emerge shouting “Bite! Bite!

Bite!” and, in some cases, actually participate in a pig-hunt. The ritual ends with the

children (now adults) on the same platform redistributing the meat of hunted or killed

pigs. Through the rite, they have transformed from prey to predator. Furthermore,

since the Orokaiva regard pigs as similar to humans, the transformation to predator is

akin to the the transformation into a warrior. This substitutability of pigs into humans

also has intimations for enemies. Moreover, in the close of the ritual, there are also

political ramifications involved in the redistribution of the meat to others from the

platform. Here, the children become adults by participating in the alliance creation and

exchange of adults. The ritual sacralizes organization into hunters (and warriors).

Reading this through Wolf’s (1990) framework, the implication is that initially the

children were just prey––insufficient in power (i) to withstand the attacks of predators

scavenged or hunted meat. Toss the wolf a leg from the deer kill––sacrifice it––and they could safely take away the rest. Early hunter-gatherer gods in many societies took the form of predators (Luckert 1975).

–– 47 ––

(ii). Through this ritual they come together and transform into hunters. Victor Turner

(1969) conveyed the importance of communitas created during the liminal stage of

transformation involved in rituals––that mutable interim between their status as

children/prey into adults/hunters––bonds between those that experienced the ritual

together, tightening that social unit. Such rituals help to strengthen the ties of alliance.

These are rituals that maintain organizational power (iii).

The ritual described here seems to follow how ontogeny recapitulates phylogeny,

whereby children go through a ritual to become adults, just as humans were originally

prey to beasts but then became predators themselves. With the concern for the

beginning of warfare, humans become prey yet again, but to other humans, which

required yet another transformation––this time of hunters into warriors (Ehrenreich

1997).

Otterbein (1989, 2004), as noted above, described the semimilitaristic

organization of hunters, their experience with weapons, and the general camaraderie of

the fraternal interest group. Big-game hunters would necessarily have to range distantly

over the territories of the big game they preyed upon. This practice would make big-

game hunters more likely to encounter other big-game hunting groups. This, he argued,

is why big-game hunters were the type of early hunter-gatherers that were more prone

to engage in warfare.

So, one could imagine a scenario similar to the early hominid practice of sacrifice:

a group kills a mammoth but is within sight of another group of hunters, who approach

as other predators. They could sacrifice a portion of the mammoth to them as to a

predator, or fight them. Given Otterbein’s (2004) arguments for warfare’s early

occurrence, perhaps the latter was chosen more often. Sacrificing or sharing a portion of

the mammoth would have entailed creation of relationships with those other hunters.

A major point to take from these theorists (Otterbein 2004; Ehrenreich 1997;

Luckert 1990) is that there were practices already in place among big-game hunters that

facilitated a transformation into warriors. That is, the cultural practices and means, or

–– 48 ––

traditions, helped to provide conditions for warfare. Secondly, it is important, according

to Otterbein (2004) and Kelly (2000), to consider the nature of social organization to

understand the conditions for warfare.

For Otterbein (2004), the second period when warfare proliferates is in keeping

with Kelly (2000) and Keeley (1996), around 10,000 BP with the onset of sedentism and

domestication. It is the creation of surplus that creates conditions of war: “The

comparative reliability of agriculture as a mode of subsistence thus transforms the

character, frequency, extent, and distribution of warfare within regional systems” (Kelly

2000:135). Similarly, Otterbein (2004: 91) argued that while there would have been

inequities among hunter-gatherers, their nomadism prevented elaboration of those

differences. Among settled agriculturalists, such differences would begin to also be

marked materially, fomenting tensions within and among groups. Chiefdoms and early

states also exhibit increasingly elaborate forms of segmentary societies.

The Concern for Causes

In anthropological treatments of warfare, it has been common to find

assessments that attribute the causes of warfare among non-state societies to a variety of

reasons, predominantly including biology, resources, social status, and territory.

According to several researchers (Thorpe 2003; Ferguson 2001; McCauley 1990), these

treatments fall into three primary camps: biological, materialist, and cultural. Since

Hobbes, many have argued that aggression is simply an innate trait of humans; in the

1960s, Lorenz (1966) and Ardrey (1966) provided a renewed popularization for such

views. Chagnon (1968) incorporated their views into a wider sociobiological argument.

Chagnon (1977, 1990) argued that there was a “reproductive striving” among

Yanomamo warriors who take women from other villages in violent raids. Warfare

allowed them to further perpetuate their own offspring in a model of inclusive fitness as

these warriors played out the survival of the fittest. In a broader yet complementary

manner, Durham (1976) analyzed warfare as collective actions that are adaptive for all

–– 49 ––

cooperating individuals. Although he was trying to unite cultural and biological modes

of explanation, his argument is underlined with a biological and materialist basis,

stating that “individuals maximize their survival and reproduction by living in social

groups and participating in collective aggression when access to scarce resources is at

stake” (Durham 1976:385). Indeed, after studying many texts of warfare for the

Northwest Coast, MacDonald (1984) noted that “first and foremost” the reason for war

was to acquire surplus food among the Haida, Tsimshian, and Tlingit.

Materialist approaches emphasize ecological conditions as the catalytic setting

for warfare. Like Durham’s (1976) “resource competition,” Ferguson (1984) had

maintained that war is waged to control resources. Countering Chagnon, Ferguson

(2001) argued that much of Yanomamo warfare was conducted over more immediate

reasons than some long-range or ultimate desire for reproductive fitness. Rather it was

for items of steel and other trade goods that were worthy of war––things that had been

introduced two decades before Chagnon arrived. Other theorists emphasized other

related aspects of resource competition. For instance, Carneiro (1970) highlighted the

importance of environmental circumscription, which then forces competition for

resources. Warfare, for Carneiro, was also the primary factor in the formation of

cultural complexity, leading to the emergence of states.

Cultural approaches to warfare involve numerous vantage points ranging from

prestige and revenge to symbolic and historical treatments. Robarchek (1989:903)

argued that most materialist and biological arguments are “ratomorphic,” borrowing the

term from Koestler (1967), meaning that their models exhibited a “variety of approaches

that continue lopping off human heads by denying relevance, if not reality, to human

consciousness, values, purposes, and intentions.” Robarchek (1989:903) provided the

Semai as an ethnographic example of how both peace and violence are purposive

actions of individuals responding to “culturally constituted” worlds.

In a similar though more cognitive approach, Harrison (1989) argued that

violence among the Avatip people of the Sepik River of New Guinea was ritually

–– 50 ––

controlled and implemented during times of social tension. However, these ritual

practices were in contrast to normal social behavior, which was regarded as peaceful

sociality. Warfare called for masks and ritual spirit-possession to temporarily transform

into warriors from their normal selves. Still, transformations into warriors were

necessary, however, this showcased that the normal conditions are commonly pacific, a

view that counters any intimations that violence and aggression are biologically innate;

they needed masks to transform from human to warrior.

In an account of warfare in northern South America, Elsa Redmond (1994)

provided a fascinating discussion of warfare in tribal in comparison to chiefdom

societies. She stated that warfare in tribal societies such as the Jivaro and Yanomamo

was primarily conducted for revenge and prestige and only secondarily for looting and

women as wives (while chiefdoms pursued territory and resources). She stressed that

such endeavors were political actions by those leading or composing an offense or

defense. For instance, the leader of a raid calls on others and persuades them to

participate in the raid:

Those individuals interested in mounting a tribal raid in order to avenge a kinsman’s death face the often long and arduous task of persuading other villagers and allies to participate.... [It] involves door-to-door canvassing, complete with formal declarations of war, rhetorical arm-twisting, and promises of war spoils (Redmond 1994:45).

Though most theories tend to fall into one of these camps of biological,

materialist, or cultural, these do not need to be so narrowly reduced. In his assessment

of the anthropological and archaeological theories of warfare, Thorpe (2003) challenged

any single theory primarily for their implications of uniformity. Rather, any theory

should be particular to its local and historical contexts. Others have emphasized an

approach that unifies both materialist and cultural in their frameworks in a theory of

practice.

Practice theorists like Bourdieu (1977, 1990) focus on the interrelation between

human agency and systemic or historical structures, which proceeds in the manner of

the dialectic; for Marx (1970 [1945]), this was the focus of his materialist approach, in the

–– 51 ––

practices of social life not in the inert objects, or materials in themselves. In archaeology,

practice theory has been discussed by Brumfiel (1994a), who stressed the agency of

methodological individualism as part of an emphasis on factional competition, and by

Pauketat (2001), who stressed historical processes in the formation of cultural practices

in Mississippian Cahokia.22 In summary for our purposes, practice theory at its core

emphasizes a situational context that accounts for environmental conditions (field),

cultural traditions and dispositions (habitus), and historical factors (structure) that

influence human practices (agency). It provides a way to discuss human warfare with its

complicated reasons and motivations while also grounding the discussion in a

materialist way without being reductionistic. As Maschner (1997) noted about

Northwest Coast warfare, “wars were never fought for a single reason.” Similarly,

Prince (2004) argued that the causes for warfare should be considered “case dependent.”

Another instance of the manifold rationales for war comes from Swadesh’s (1948)

discussion of the motivations of Nuu-chah-nulth (formerly Nootka) warfare. He

analyzed nine war narratives collected by Edward Sapir that contained forty-five

accounts of wars. He found the motivation to fight for territory to be more important

than plunder. One example has the Ucluelet people holding a war council because “The

Ucluelet had no river....” After visiting and being feasted at different villages in the

greater area, they decided that the Namint controlled a river with the best salmon. “’All

right,...’ they said. ‘Let us kill them off and take away their river.’” Everyone in the tribe

consented (Swadesh 1948:84-85). However, Swadesh (1948:85) also noted that in

general, “the food supply on Vancouver Island was generally very plentiful, we must

not assume that the desire for new territory was typically based on a real shortage of

food.” Here, Swadesh indicates that while territory was an aim sought in warfare, it

was not for food in and of itself; there were other driving factors. He also found

22. Also, McGuire’s (1992:252-47) Marxist archaeological approach contains a theory of praxis that stresses the dialectic not only between agents and structures, but also the dialectic between researchers and their practice.

–– 52 ––

instances where the motives were for increased rank, for capturing slaves, gaining

plunder, and seeking revenge. In fact, Swadesh found it difficult to make clear

distinctions among these, not causes, but “motivations”:

Retaliation and hope of gain have to be seen together. In almost every instance, those who propose warfare are able to point to some offense that has been committed against them; on one occasion ... the offense is invented. In some cases, warfare is proposed to the tribe simply and frankly in terms of advantages to be got. Sometimes revenge and gain are mentioned together. In a few instances, the narrative indicates clearly that retaliation is a mere pretext or at least a very secondary consideration. Even in the many passages in which no other motive than retaliation is mentioned, we can only assume that those who sat in council and planned the raids must have been very sensible of the specific material benefits and the prestige that would come to them from a successful conclusion of the project (Swadesh 1948:91).

Instead of trying to reduce the multifaceted reasons for warfare to the

procrustean bed of one cause, I find it more effective to use the exchangeability of

capital, from practice theory, that can incorporate multiple motivations into its larger

frame, as Swadesh had found.

Anarchist Theories of Warfare

Yet for all their limitations, they [non-state] societies show that the Hobbesian nightmare of universal war in a “state of nature” is a myth. A society without hierarchy in the form of rulers and leaders is not a utopian dream but an integral part of collective human experience.––Peter Marshall (1993:13)

Marshall (1993) pointed out that the notion that a society without rulers is often

seen as constant war. However, as many anthropologists of war have noted (e.g.,

Otterbein [2004], Kelly [2000], and Keeley [1996]), warfare is more frequent in later

periods of human evolution, when the scale of social organization increases, diversifying

into segmentary societies that develop opposed interests. Accordingly, most of human

existence since the onset of Homo sapiens has been one without warfare (if not without

violence), even when these societies had no rulers, or were anarchic. Given this context,

it appears as if this Hobbesian notion is itself an ideology of states––that it is in the

–– 53 ––

interest of rulers to perpetuate a notion that society would erupt in chaotic warfare if

their rule ended. Indeed, it is with state-scale societies that warfare is more common

and manifests in its deepest manners of destruction. As Randolph Bourne (1999 [1918])

had argued, “War is the health of the state.”

Marxists and anarchists do have some agreement in being not so opposed to war,

or at least certain types of war. Marx and Engels tended to favour the importance of

economics, but they did not overlook the effects of warfare in history. In fact, Marx and

Engels considered warfare a major driving force for much of human history, particularly

for early states (Hobsbawm 1964:44-45). Not inclined to pacifism, warfare was

important for them for revolution, to ultimately achieve communism. Lenin, referencing

Clausewitz’s (1911) famous phrase, stated that “War is not only a continuation of

politics, it is the epitome of politics” (Kiernan 1983:522). Marx and Engels also viewed

national armies somewhat positively at first, thinking that these would help train the

proletariat for the time of revolution. However, they came to see wars as ends in

themselves, while Lenin blamed capitalism for war (Kiernan 1983:522-523).

Anarchists have a negative connotation partly through their association with

violent acts in the early 19th century, with several assassinations of many heads of state

(e.g., Tuchman 1966:63-116). Bakunin (1842) was infamous for saying that the “The urge

to destroy is also a creative urge.” But, that is only part of the spectrum, as many strains

of anarchists have been or are stridently pacifist, particularly Tolstoy (1990). A

commonality, however, is that warfare is warranted in countering unjust rule. Even

Kropotkin, a longtime pacifist, was heavily criticized for supporting the efforts against

Germany in World War I, as he thought it ultimately furthered their cause (Miller

1976:225-232). He also supported the Russian Revolution while countering its eventual

direction, ultimately becoming a critic of Lenin. The point is that, for anarchists and

Marxists, warfare has not been seen as something necessarily to be viewed as negative.

In fact, acts of war actually can be a very just practice and are warranted to counter

tyrants and bring about greater freedom for individuals.

–– 54 ––

Pierre-Joseph Proudhon (1861) wrote a two-volume treatise on warfare, called La

Guerre et la Paix, or War and Peace––which is often maintained as the source for the title

of Tolstoy’s novel. Proudhon aimed to counter the common notion among his fellow

French philosophers that war was a crime against humanity, a pathological social

practice (Noland 1970). Instead, he argued warfare was conducted by those who did not

accept the circumstances as given, by those who would not be passive in the face of

threats to their freedom. Proudhon (1861) viewed warfare as an enaction of human

agency par excellence: “War, in one form or another, is essential to our humanity.... [It]

is inherent in humanity.... [It is] the most grandiose manifestation of our individual and

social life” (Noland 1970:290).23 In this sense, it is an agential assertion of autonomy. As

Žižek (2003:31) has stated, similar to Proudhon, assertions of autonomy necessarily have

to include the option of violence:

... [A]uthentic revolutionary liberation is much more directly identified with violence––it is violence as such (the violent gesture of discarding, of establishing a difference, of drawing a line of separation) which liberates. Freedom is not a blissfully neutral state of harmony and balance, but the very violent act which disturbs this balance.

More than that, war was, Proudhon (1861) argued, “essentially justicière and

juridique,” or related to justice and law. As Noland (1970:290) noted, “it [warfare,

according to Proudhon] has been an agency through which justice has found expression

and affirmation in the world. For this reason ‘war is legitimate in its essence, saintly and

sacred ...heroic and divine ...the summit of human virtue.’” Proudhon viewed war as

the “most ancient expression of justice in society,” that it had a “moral and juridical

core” (Noland 1970:299).

In Proudhon’s (1861) view, warfare is simply an outgrowth of normal tensions

and conflicts within human societies. Societies are not stable in any sense, but dense

with oppositions and tensions. A society must be viewed as perpetually within a

23. All quotes from Proudhon’s (1861) La Guerre et la Paix, in this section are drawn from Noland’s (1970) translations summarizing Proudhon’s sociology of war; the full two-volume work is not yet translated into English.

–– 55 ––

Heraclitean state of becoming in which “everything relates to everything else, is linked

up with it; consequently, everything is in mutual opposition, balance, and equilibrium.”

This is the “eternal dance,” where the lives of individuals are “une solution d'antinomies

sans fins,” or the resolution of antinomies without end, a never-ending dialectic of life.

These internal tensions manifest between individuals and their societies at many scales,

yet this did not mean there was chaos, the Hobbesian war of each against all:

While man was an animal, he was, Proudhon insisted, following Aristotle, an animal of a particular sort––a “social animal,” whose destiny it was to live in society. On the one hand each individual was “at war” with his fellow men. At the same time he was “moved by an internal attraction to other individuals”––“moved by a secret sympathy” which he could not resist without denying “his own nature,” for man’s “social needs” were “complex and imperative” (Noland 1970:293).

Here is a common point by anarchist theorists that there is tension, even conflict,

at all scales of society from the individual, to their families, to broader society, and those

beyond. However, this is not an atomistic portrayal of individuals, rather, it is an

autonomous one, where individuals also seek commonality and shared interests with

others. It should be noted, that this is opposed to those sociobiological portrayals of

humans as self-interested “agents,” where every instance of aid, selflessness, or altruism

is somehow reduced to pure self-interest. Instead, humans are first “social animals” that

desire sociality and to help others in mutual aid, where individuals willingly curtail

some of their own freedoms to live and accomplish projects with others.

While praising many instances of warfare, Proudhon (1861) also recognized that

warfare, as conducted by states, was usually not in the cause of justice and increased

freedoms for those involved. Instead, he posited that the social inequity of states leads

to poverty among the majority and a tyranny of the minority as they attempt to maintain

their position, manifesting in corruption and violence. Bourne (1918) argued that states

pursue war, not only for material interest, but to also settle the internal tensions caused

by inequities and direct those angers elsewhere, towards others, achieving the

“peacefulness of being at war.”

War is the health of the state. It automatically sets in motion throughout –– 56 ––

society those irresistible forces for uniformity, for passionate cooperation with the Government in coercing into obedience the minority groups and individuals which lack the larger herd sense (Bourne 1918).

States, as Jonathan Friedman (1998, 2003) argued, attempt to engender

uniformity through hegemonic actions. Ideologically, political rituals are often effective

in unifying a populace even when there may be great disparities amongst those who

witness the ritual. As Kertzer (1988) well argued, there is a multivocality to political

rituals, when performed well, that allows for popular agreement on the symbol or ritual

despite deep disagreements over what those symbols and acts mean––a powerful device

for politicians to employ. These types of political rituals can be used to tighten support

for a war.

Bourne (1918) noted that in recent history wars are the actions of states, not

societies or nations of peoples––it is always declared by those in power: “There is no

case known in modern times of the people being consulted in the initiation of a war.”

Accordingly, wars are not voted on; they are declared by the leaders of states. While

many states use arguments of defense, Proudhon (1861) found most state wars to be “a

frightful caricature of the forms of justice.” This is consistent with the anarchist theory

of history, described above, wherein those in power aim to maintain their power, which

is a top-down approach to a society. In such scenarios, anarchists have argued, it is

warranted to challenge such authoritarians for their misuse of power.

Kropotkin (1946 [1897]) similarly found this notable difference between wars of

small-scale societies and those of states:

[In communal societies] the struggle was for the conquest and defence of the liberty of the individual, for the federative principle for the right to unite and to act; whereas the States’ wars had as their objective the destruction of these liberties, the submission of the individual, the annihilation of the free contract, and the uniting of men in a universal slavery to king, judge and priest––to the State.

The rise of the state, for many anarchists, is the major development in

humanity––not domestication, agriculture, urbanization, nor industrialization. The state

is the great “rupture” ("coupure") as Pierre Clastres (1987:202) termed it, as it is not a

–– 57 ––

revolution fomented by a new economic technology, but a political revolution in the

organization and concentration of power in society; in Wolf’s (1990) framework, the

concentration of political power in a state is the institutionalization of structural power

(iv). Clastres’ (1987, 1994) ethnographic work focused on “stateless societies,”

describing those, not as societies in which states have not yet developed, but rather as

ones that have resisted forms of state power and its associated inequality. Gledhill

(1994) argued that Clastres’ perspective consisted of a variant of Sahlins’ original

affluent society in political rather than economic form. Warfare, among small-scale

societies, served to prevent such concentrations of power:

The atomization of the tribal universe is unquestionably an effective means of preventing the establishment of socio-political groupings that would incorporate the local groups and, beyond that, a means of preventing the emergence of the State, which is a unifier by nature (Clastres 1987:213).

As emphasized by Kropotkin (1946 [1897]), the state is not equivalent to society,

government, nation, or country. Rather, the state is a political formation, which tries to

ideologically mesh its conception with that of nation and country. This distinction

between the two is expressed in the title of Clastres’ (1987) work, Society Against the

State.

Interrelating Scales through Alliance Formation

In his summary analysis of theories of warfare, Helbling (2006) emphasized that

the anthropology of warfare could benefit from theorists of international relations and

foreign policy, who use the notion of “anarchy” to describe situations of warfare, as

described above; a few other anthropologists have also started to use the same

terminology, such as Snyder (2002) and Schon (2008). This use of “anarchy” derives

from its connotation of chaos under a lack of rulership and is not associated with the

theory of anarchism, which is about a form of social order. Here, I argue that this use of

anarchy is rather simplistic––it simply means the polities in warfare act autonomously.

Anarchism as a theory includes autonomy but provides a much larger framework

–– 58 ––

within which to assess such interactions. With this qualification, Helbling’s (2006:213)

analysis provided an interesting insight, which is that previous theorists pondering the

causes of warfare have failed to account for the formation of alliances:

The phenomenon of alliance has been neglected by anthropology so far, but this must be taken into consideration because whoever has to wage war also needs allies. Alliance formation influences the regional relation of force between warring local groups, and both victory and defeat may depend on the support of allies (Helbling 2006:213).

This is why Helbling considers alliances a “crucial phenomenon” in the context

of war. In Wolf’s (1990) categorization, alliances represent an aspect of organizational

power (iii), and so, arguably alliances are implied if not overt in the works of Otterbein

(2004) and Kelly (2000) which highlight the organizational context of war. Still, Helbling

is pointing to an important aspect of organization in warfare, particularly as his

description of alliances is not restricted to alliances of kin groups.

There are certain dynamics to alliances that should be considered. For one, the

creation of an alliance can drastically alter the nature of power relations in a social field.

As Helbling (2006:125) defined, “Alliances are forms of pragmatic co-operation based on

(short term) common interests: two groups will do better against an enemy group by

forming an alliance.” Thereby, the strongest group can easily lose its position of

dominance when two (or more) weaker groups form an alliance.

An important aspect of incorporating alliance theory into a considering of social

organization is that it accounts for internal tensions within a society. Antonio Gilman

(1984) stressed tensions exist within society alongside expressions of solidarity. He was

countering Durkheim’s (1949) conception of social solidarity, which he considered

inadequate because it did not account for internal tensions. Gilman argued, using

alliance theory, that people form coalitions with subsets of the overall society for

security reasons: if resources are stretched too thin for all, not everyone will survive;

however, if you limit your social obligations to a subset, with “closed connubia,” you

can better ensure security at least within your coalition. He noted that the distinction

was minor, but led to an entirely different social dynamic that could involve significant

–– 59 ––

qualitative changes in the archaeological record. Although Gilman (1984) advocated this

as a Marxist approach, it differs from classic Marxism in that it does not focus on class-

based tensions, which would be centralized or vertical. Instead, he emphasized allied

groups opposed to other groups, which is decentralized or horizontal; that is, his

argument has parallels with an anarchist argument.

Alliance theory has also been advanced by the development of factional

competition (Brumfiel and Fox 1994). In her introduction, Brumfiel (1994) argued that a

factional-competition approach countered the ineffectiveness of the two predominant

approaches, cultural ecology and Marxism. Building upon Barth (1959), Bailey (1969),

and Giddens (1979), she proposed a practice theory with influences from methodological

individualism that emphasized the actions and agency of individuals that have to work

within the structures of history and culture. Factions are groups, assembled with self-

interested members by resourceful leaders, that compete with other groups for prestige

and power. Within the volume, contributers described various mechanisms for creating

such coalitions: feasting, warfare, marriage, ethnic identity, and so on. Two main

principles involve alliance building and competition, both of which leave archaeological

traces. Clark and Blake (1994) referred to these factional leaders as “aggrandizers” and

argued for the use of elaborate feasts to cohere allies in the Mazatan region in Chiapas.

They noted importantly that indicators of rank appeared prior to the rise in population

growth, countering cultural ecological arguments. Redmond (1994) also highlighted the

importance of alliance formations, which was critical to the success of chiefs among the

constantly feuding Jivaro. These studies of factional competition, argued Brumfiel

(1994), countered the emphasis of Marxism on class-based antagonisms. Class-based

analyses were insufficient and inappropriate for groups where internal strifes and

alliance formations do not work in class-based fashion but rather among factional lines;

this of course, can also be seen as having affinities with anarchist critiques of Marxism.

Not only do societies contain multiple internal tensions, but so do alliances.

Members strive not only to be in the stronger coalition but also to gain the most within

–– 60 ––

the coalition itself (Helbling 2006:125). Elsa Redmond (1994) emphasized the

importance of the promises of spoils for alliances formed for raids on other Jivaro

groups. If the gains from the spoils do not meet their satisfaction, the alliance can

dissolve rather quickly.

One criticism I should note about these coalition theories is that these are based

upon common economic conceptions of individuals as ego-centered agents where

individuals could not convincingly commit an act of altruism if they tried. Anarchism is

not based in alliances simply for advancing self-interested goals––although that element

is present. Anarchism instead also allows for other reasons that people might want to

work together in mutual aid: that allied work on big projects might be easier, that

individuals are social beings, and they want to form close bonds with others. As Studs

Terkel (2005) has observed:

And this is my belief, too: that it’s the community in action that accomplishes more than any individual does, no matter how strong he may be.Einstein once observed that Westerners have a feeling the individual loses his freedom if he joins, say, a union or any group. Precisely the opposite’s the case. The individual discovers his strength as an individual because he has, along the way, discovered others share his feelings––he is not alone, and thus a community is formed (Terkel 2005).

In his Theory of Political Coalitions, Riker (1962, quoted in Wang 2005:1) pointed

out, “in the three-or-more-person game, the main activity of the players is to select not

only strategies, but partners. Partners, once they become such, then select a strategy.”

The implication here is significant: one chooses the practices one will employ, but also

one’s allies; the alliance itself subsequently determines the practices that will be used.

Moreover, Riker’s formation suggests a level of agency on the part of coalitions, an

agency that is still anchored in the willing participation of individuals who compose the

alliance––therefore, it is not something to theoretically reify as an actor or agent in its

own right, although it is understandable how the effect of such coalitions could be

interpreted that way. Coalitions are a gestalt of individual human agency.

Since Riker’s (1962) major study, some have formed typologies of coalitions.

–– 61 ––

Wang (2005) provided a descriptive and historical typology of coalitions, categorized by

the social formation, duration, task, and status. Each type of coalition, he noted, is a

political formation: “Since coalescing partners cooperate with only some of the other

actors and struggle against the rest of the actors, any coalition is in essence seeking

influence directly among actors without mediation of materials and therefore is

political” (Wang 2005:1). Social coalitions involve the types of groups that ally because of

similar identity or shared topical interest (religious, ethnic, familial, etc.). Duration

coalitions are defined by the length of the alliance: short-term, middle-term or long-term.

Task-oriented coalitions involve alliances oriented towards single goals; accordingly, these

are mostly short-term. Status coalitions involve alliances categorized by combinations of

small- and large-scale, high and low position, or strong and weak constituents; status

refers to status of one group relative to the other, not status as prestige. While only

typology, it is a framework to begin to understand further aspects of how coalitions

form, enlarge, strengthen or dissolve.

Amanda Tattersall (2006) provided a study of alliances derived from organizers

of unions and activists. She understands that the nature of the coalition will influence its

effectiveness, power, and longevity. She classifies coalitions into four primary categories

beginning with the simplest and most temporary: ad hoc, support, mutual-support, and

deep coalitions (Table 1). The simplest formation involves ad hoc coalitions that may be

event-based or task-oriented. Once the task or event is over, the coalition is also over.

Typically, these are “fleeting” and may involve minimal amounts of alliance from the

heads of the groups, with not much interaction amongst the respective supporters or

members. These are tactical alliances only, and not strategic. While there is a common

interest, there is no “joint decision making.” Ad hoc coalitions can sow the seeds,

however, for more permanent alliances. Support coalitions are short-term, but structured

relationships between groups. They allow for “closer organisational connection through

joint decision making” (Tattersall 2006:4). The weakness of support coalitions are that,

while they are able to rally people quickly (as a “rent-a-crowd”), there is little

–– 62 ––

Table 1: Types of coalitions (adapted from Tattersall 2006) .

Type/Trait Ad Hoc Support Mutual-Support Deep Coalition

Interest An event/issue Specific to one group in coalition

Common interest of all in coalition (particularly the leaders)

Common interest (andideology) of all in coalition (members and leaders)

Duration Episodic Short-term Short to middle term Long term

Organization Loose Some shared decisionmaking (usually led bygroup with main interest)

Shared decision making (connections through leaders from member groups)

Decentralized structure (connectionsthrough individuals of member groups)

organizational power or “buy-in” amongst those involved due to their short-term nature

and structure. Instead, stronger coalitions are able to accomplish that. Mutual-support

coalitions form around more than one issue or event, tying the bonds of the coalitions

into a stronger alliance that can be seen as more than temporary. This broadens the

shared interests and creates the need for closer strategy between those allied. Finally,

deep coalitions tighten the bonds of a mutual-support coalition into something that is not

just coordinated by group leaders but also by the interaction amongst the members of

allied groups, facilitating decision-making at a variety of scales. The alliance is no

longer mainly by the groups’ leaders and becomes a “decentralized structure” (Tattersall

2006:12) with multiple types of and scales of capacities for mobilization. Through her

typology, Tattersall was able to argue that coalitions are more effective through

increasing common interests, decision-making abilities, and deepening involvement

between allied groups. As she noted, “the power and capacity of a coalition is greatly

influenced by its internal practice.” These coalition typologies may have some utility for

considering coalitions archaeologically, as Helbling (2006) has done.

While Helbling (2006) did advance the anthropology of war by emphasizing

alliances, his treatment of alliances is limited in one respect. He is accurate to note, for

instance, that “no serious wars will break out as long as two adversaries are of about

even strength and in a stalemate situation.” However, the context for such claims are

limited only to the actions of those two groups. The actual social field of most groups –– 63 ––

involves not only one other group, despite how it may dominate their consciousness at

any one time, but rather needs to include all possible groups. The nature of opposition

between groups may be conditional. Groups are known to ally even with local enemies

to face larger external threats. That is the nature of alliance formation––it is not limited

to one scale, but needs to incorporate several social scales. Otherwise, Helbling’s (2006)

model works well for any particular scale: family, household, household cluster, village,

village cluster, tribe, region, nation, and so on. Had Helbling used the theory of

anarchism, rather than “anarchy” after political scientists’ use, he may have more

readily envisioned this. Anarchists have theorized the development of alliances and

their nature from the local group to larger scales of social action.

A model for the creation of anarchist organization from the small-scale or local to

larger-scale organizations is seen in late 1930s-era Spain with the FAI, or Federación

Anarquista Ibérica. They set up grupos de afinidad or “affinity groups” to help foment the

revolutionary spirit to fight Franco and the fascists. Marxist groups were also involved

in the revolution, however, they argued over the organizational forms, with the Marxists

advocating a centralized structure and the anarchists favouring a decentralized one:

The Marxists argued that their organizational forms gave them greater efficiency and effectiveness, a claim the Anarchists emphatically denied. To the contrary, they insisted that the most efficient and effective organization was ultimately based on voluntarism, not on coercion or formal obedience. A movement that sought to promote a liberatory revolution had to develop liberatory and revolutionary forms. This meant ... that it had to mirror the free society it was trying to achieve, not the repressive one it was trying to overthrow. If a movement sought to achieve a world united by solidarity and mutual aid, it had to be guided by these precepts; if it sought a decentralized, stateless, nonauthoritarian society, it had to be structured in accordance with these goals. With voluntaristic aims in mind, the Anarchists tried to build an organic movement in which individuals were drawn to each other by a sense of “affinity,” by like interests and proclivities, not held together by bureaucratic tendons and ideological abstractions. And just as individual revolutionaries were drawn together into groups freely, by “affinity,” so too the individual groups federated by voluntary agreement, never impairing the exercise of initiative and independence of will (Bookchin 1998:108).

Bookchin (1971:243) commented that these affinity groups “could be regarded as

a new type of extended family, in which kinship ties are replaced with by deeply

–– 64 ––

empathetic human relationships––relationships nourished by common revolutionary

ideas and practice.” When they held congresses of these groups, they were titled

“asembleas de las tribus,” or assemblies of the tribes (Bookchin 1971:243). While they did

confederate into local, regional, or national forms, the emphasis of power remained with

the local, affinity group: “The groups proliferate on a molecular level and they have

their own ‘Brownian movement.’ Whether they link together or separate is determined

by living situations, not by bureaucratic fiat from a distant center” (Bookchin 1971:244).

The important point Bookchin is making is that alliances are temporary and conditional.

Like the anarchist view of authority, the power alloted to the alliance must be seen as

justified by its need, needs that can shift with changes in conditions or historical

circumstance. As power is centered at the smallest scale, in the extended family or the

affinity group, the need or justification for large-scale alliances would be much weaker

than alliances closer to those that affect the daily routines or concerns of local groups.

Large-scale alliances would, accordingly, be more temporary and event- or task-oriented

than alliances at lesser scales.

Conclusion

Many anthropological treatments of warfare have tried to reduce its occurrence

to one primary cause, rather than viewing such interactions as the complex conjunction

of multiple factors. Manifold reasons can be employed as causes or justifications for

warfare. In chess, one could view his or her next move confronting various needs: to

defend a threatened piece, to mobilize another, or to attack an opponent’s piece. But, a

move could be chosen that satisfies all three requirements at once. In analysis of it, one

could not reduce that action to one causal motive or reason, rather, the single move

encapsulates all three needs, and the better player employs such moves. Similarly, in

times of war, a leader can deploy actions that have multiple effects at once––for instance,

performing one military act that furthers their material gain, builds the strength of one’s

defense, and satisfies allies politically or ideologically. In dealing with complex

–– 65 ––

societies, or other human societies, we should avoid such treatments that are

reductionistic, that do not recognize the multifunctionality of single actions, formations,

or tools, or fail to incorporate the multiscalar context within which these practices are

performed.

While overall studies such as Otterbein (2004), Kelly (2000), and Keeley (1996) are

intriguing, the findings are on a scale that is broad. The generalities rarely match

particular cases, especially when they involve the complex hunter-gatherers of the

Northwest Coast, who often are described as exceptions to general anthropological

categories for foragers. More useful are specific case studies that tackle local conditions

and historical circumstances. Several volumes have followed upon those works

recently, including Rice and Leblanc (2001) for the U.S. Southwest; Chacon and

Mendoza (2007a, 2007b) and Chacon and Dye (2007) for North and South American

studies; and Arkush and Allen (2006) and Otto, Thrane, and Vandkilde (2006) with

worldwide cases and perspectives.

What makes the scale and perspectives of Otterbein’s (2004) and Kelly’s (2000)

works so different from those that advocate single causes for war is that they both argue

for the importance of social organization in the manifestation of warfare. They

emphasize its sociality and do not reduce warfare to factors outside of human culture.

There may be population pressures, climatic changes, or environmental constraints, but

they recognize that the ultimate decision to go to war is a social one. What’s clear is that

warfare is a practice––it is an option that groups can pursue, however, there are others

as well that do not involve violence. Groups can negotiate to share resources; they can

intensify their use of other resources in the face of drought or flooding, or seek aid from

extended family and allies. The point is, warfare is not the direct result of such

“causes”––the use of “cause” is a problematic term here, when discussing warfare

among human groups. Rather, what should be discussed, especially among

anthropologists, are the reasons or rationales for warfare, or as Swadesh (1948)

described, “motivations.” Causes, or causation, has connotations of physics, with

–– 66 ––

images of unidirectional objects impinging on another on a grand billiards table.

Reasons can be interchangeable with causes, however, its origins are from the Latin,

ratio, which is from the verb reri, “to consider.” Instead of the cause-effect impacts of

billiard balls, the actions of humans in warfare are considered––just as chess-players

consider the movement of pieces on the chessboard, contemplating their options,

deliberating over the opponent’s motivations, and responding to the changing

conditions of the board for each turn of events in a simulacrum of war.

Rather than reducing warfare to biology or to limits of carrying capacities, an

approach is needed that can be explanatory at multiple scales, ranging from the

individual to larger societal groupings. Practice theory, with its Baileyan (1969)

characterization of the competition by individuals for the spoils, provides a framework

for assessing motives that are not reduced to a singular meaning. Indeed, with the

exchangeability of capital––where resources or loot acquired in warfare can be traded

for enhancing one’s symbolic status, for example––many reasons can be seen to be

operational at once. Moreover, with numerous warriors in battle, there can be many

various reasons for each individual to participate. At a larger scale, the theory of

anarchism provides a scale of analysis with principles of organization for those cultural

practices; anarchism emphasizes a social rather than individual scale, involving

principles of decentralization and mutual aid through networked alliances. These are

principles and practices that link individuals into larger units of action.

I have tried to show how the framework involving anarchism, practice, and

power is useful for considering the anthropology and archaeology of warfare. It does

not replace other theories discussed above, rather, I argue that it provides a framework

to interconnect disparate theories of warfare at a variety of scales. It provides an

overarching frame that assesses the power and capacity of warfare––in which

organization is seen as a criterion for warfare as opposed to violence according to

Otterbein (2004) and Kelly (2000). Wolf’s (1990) modes of power incorporate their

emphasis on organization (iii) by also readily including the dimensions of power of

–– 67 ––

individuals (i) and between individuals (ii), while also making a distinction for a higher

dimension of structural power (iv). Indeed, Otterbein’s (2004) classification of two major

periods of warfare parallel these: warfare amongst hunter-gatherers as organizational

power (iii) formed in fraternal interest groups, which engage in a constant yet not

unifying mode that serves ultimately to redistribute power. Early state warfare could be

seen as a structural reformulation of warfare, altering the nature of warfare into

something that facilitated the concentration of power rather than redistributing it.

Practice theory aids in helping to understand the nature of that organizational

formation through alliance building, which is seen as crucial and under-considered by

Helbling (2006) for an anthropology of warfare. Individuals and groups do not only

improvise among a set of practices tactically and strategically to defeat their enemies in

warfare. They also politically act to acquire and maintain allies as partners, knowing

that a larger-scale coalition is the most effective offense, particularly when technology

and defenses are the same as the opponents’. Practice theory also provides a meta-

framework for viewing long-standing debates over the causes of warfare, which too

often try to reduce these complex issues to single factors. A perspective of the

exchangeability of capital allows more readily for the inclusion of multiple causes, from

material to symbolic, as actions can serve more than one need or advance more than one

cause at once. Finally, anarchism provides a larger frame that shows how these

practices are organized. Anarchism also provides a frame to interconnect other theories

of warfare used in archaeology. Factional competition and alliance formations can be

viewed within the larger framework of anarchism, which allows for broader integration

of those disparate theories.

Now having situated our theoretical framework of power, practice, and

anarchism within the anthropology of warfare, we will now situate this inquiry within

the ethnohistoric evidence for warfare in the Coast Salish region.

–– 68 ––

Chapter IV: Histories of Warfare from Documents and Oral Histories

The first glimpses of the Coast Salish by early explorers and traders in the late

18th and early 19th centuries, provide surprisingly useful, if fragmentary, accounts of

violence and warfare in the region (Figure 2). Occasionally, these early encounters

highlight the ways in which Coast Salish peoples displayed warlike intentions,

interactions, and both offensive and defensive preparations in response to the colonial

process. It has to be stressed that these documents, while often illuminating, were

produced as part of the colonial encounter and should be considered within the

postcontact context. The reach of colonial impacts affected Northwest Coast societies in

general, and Coast Salish communities in particular, many years before our earliest

accounts. While in some cases, there are descriptions from various journals and logs

noting that they must have been the first Europeans encountered by many groups, it is

clear that their reputations preceded them, along with European trade items like iron,

and deadly diseases such as smallpox.

A period of warfare and defensive site construction occurs after contact. While

the context has been altered from how the Coast Salish operated before contact, these

glimpses of their lifeways still provide indications of how they employed traditional

methods or practices and how they co-opted new technologies and situations to their

pursuits.

Early Accounts of Warfare

The language of these natives differs much from those on the outer coast. They recognize no superior chief and carry on continual warfare with those on the north side, thus accounting for the fact that the beaches are strewn with the harpooned heads of their enemies. They are affable,

–– 69 ––

-126  -124  -122 

-120 

48 

50 

0 25 50 km

Washington, US

-124  -122 

Vancouver

IslandGulf of

Georgia

Fraser River

Fort Langley

(est. 1827)

Fort Victoria

(est. 1843)

Haro StraitJuan de Fuca Strait

Dungeness Spit

Olympic

Peninsula

Point

Grey

Point

Roberts

Shingle

Point

Desolation

Sound

Quadra

Island

Cape

Mudge

Discovery

Harbour

I-eh-nus

Whidbey

Island

Qualicum

River mouth

Maple Bay

Lamalchi Bay

Cowichan Bay

Toba

Inlet

Clallam

Bay

Port Townsend

Fort Seattle

(est. 1851)

Figure 2: Map showing locales of early encounters, forts, and most places mentioned in the chapter.

happy, of good stature and well formed, but the different kinds of paint with which they disfigure their countenances make them horrible to behold. ––Manuel Quimper Benítez del Pino (Wagner 1933:131)

In the summer of 1790, the Spanish Quimper Expedition entered the waters they

named the Juan de Fuca Strait, encountering Central Coast Salish groups including the

Klallam, Sooke, Songhees, and Saanich, as well as numerous groups of people who lived

on the San Juan Islands. In one encounter, they sailed past Dungeness Spit on the

Olympic Peninsula and made note of the “harpooned heads of their enemies,” quoted

–– 70 ––

above, that were likely impaled by the Klallam. Quimper also remarked upon the

wearing of hide for an armour, probably of elk (Wagner 1933:131).

The second Spanish Expedition headed by Francisco Eliza entered the Strait the

following summer. Captain Eliza recorded that native canoes always approached and

their occupants had bows and arrows at the ready. In the Haro Strait, Coast Salish

groups surrounded a ship on an exploration led by one of Eliza’s men, José Verdia.

Under threat, he fired his cannons at them and sank one canoe, noting that they had

“killed some natives among those who were striving to attack the long boat from all

sides with some heavy spears having points of bone like harpoons” (Gormly 1977:27).

Suttles (1989) thought these attackers were likely Saanich or Cowichan. Another crew

member named Pantoja also inscribed an account of the expedition. He described the

Coast Salish as more bellicose than the groups on the outer coast:

They seem to me, however, to be more warlike and daring, not only on account of what happened to the longboat but because from the Puerto de Quimper [New Dungeness] to the Ensenada de Rojas [Clallam Bay], some 18 leagues, we saw on all the beaches a number of skeletons fastened to poles of the shores. Ordinarily they all use some thick hides dressed like deerskins and in the bows of their canoes they carry long spears (Wagner 1933:189-90).

In 1792, another Spanish expedition, led by Galiano and Valdés, also remarked

upon an apparent disposition to warfare by the Coast Salish, specifically the Musqueam

of Point Grey:

[The Musqueam] displayed an unequalled affability together with a warlike disposition. They traveled provided with many good arms such as iron-pointed spears half a yard long, quivers full of arrows with tips of the same metal and of flint, bows and clubs, and hold the latter in such high esteem that it was not possible to get one in exchange for knives of Monterrey shells24 (Wagner 1933:260).

In his analysis of these accounts, Suttles (1989:261) pointed out that no

fortifications are mentioned, however, an artist aboard one of the Spanish expeditions

did depict one described as located on the Strait of Juan de Fuca (see Figure 48, pg. 268).

24. These “Monterrey shells” refer to abalone shells that were brought up the coast from California specifically for trade.

–– 71 ––

The 1792 Galiano and Valdés Expedition encountered not only the Coast Salish, but also

Captain George Vancouver’s Expedition. In Howe Sound, Vancouver described some

Coast Salish, likely Squamish, who approached their ship, noting face-paint and arrow

materials:

We had seen about seventeen Indians in our travels this day, who were much more painted than any we had hitherto met with. Some of their arrows were pointed with slate, the first I had seen so armed on my present visit to this coast; these they appeared to esteem very highly, and ... took much pains to guard them from injury (Vancouver 1984 [1792]:587).

Like the Spanish before them, the British also noticed the presence of defensive

sites. Menzies (1923 [1792]) described a fortified Coast Salish village near Homfrey

Channel on the mainland side of the Strait of Georgia, within Desolation sound:

At the farther end of these Islands we come to a small Cove in the bottom of which the picturesque ruins of a deserted Village placed on the summit of an elevated projecting Rock excited our curiosity and induced us to land close to it to view its structure. This Rock was inaccessible on every side except a narrow pass from the Land by means of steps that admitted only one person to ascend at a time and which seemed to be well guarded in case of an attack....

Sixteen years later, when Simon Fraser undertook his exploratory trip down the

Fraser River in 1808, he also encountered frequent indications of warfare. At one point,

he described how the “The Indians advised us not to advance any further, as the natives

of the coast or Islanders were at war with them, being very malicious, and will destroy

us” (Fraser 1960 [1808]:104). By “Islanders” he appears to have heard about the

Cowichan, who conducted raids up the Fraser River. At the Fraser’s delta, lived the

Musqueam, a group who were reportedly feared by some upriver peoples.

At last we came in sight of a gulph or bay of the sea [the Strait of Georgia]; this the Indians called Pas-hil-roe. It runs in a S.W. & N.E. direction. In this bay are several high and rocky Islands whose summits are covered with snow. On the right shore we noticed a village called by the Natives Misquiame [Musqueam]; we directed our course towards it. Our turbulent passenger conducted us up a small winding river to a small lake to the village.Here we landed and found a few old men and women, the others fled into the woods upon our approach. The fort is 1500 feet in length and 90 feet in breadth. The houses, which are constructed as those mentioned in places, are in rows; besides some that are detached. One of the natives

–– 72 ––

conducted us through all the apartments, and then desired us to go away, as otherwise the Indians would attack us (Fraser 1960 [1808]:105-106).

In continuing downriver, Fraser described his wariness regarding the Coast

Salish groups in the area: After [a] skirmish [at Musqueam] we continued untill we came opposite the second village [between Musqueam and Point Grey]. Here our curiosity incited us to go a shore; but reflecting upon the reception we experienced at the first, and the character of the Natives, it was thought neither prudent nor necessary to run any risk, particularly as we had no provisions, and saw no prospect of procuring any in that hostile quarter (Fraser 1960 [1808]:106).

Upon returning upriver, Fraser (1960 [1808]:109) commented that some groups

said that his party must have experienced “good fortune to escape the cruelty of the

Masquiamme.” Fraser did not just see defensive aspects of villages, and warlike threats,

but also saw what he described as scalps: “one of the crew had a large belt suspended

from his neck garnished with locks of human hair” (Fraser 1960 [1808]:109).

The Fort Langley Journals

Wednesday 19th [March 1828]. Clear frosty morning. Three Indians from the Kutche Camp up Pit’s River informed us that the Cowitchen war party were passed, that they killed 10 of the Penault [Pilalt] tribe and had taken a number of their women & Children Slaves. They must have passed here at night. This tribe lives about a day’s march up. This warfare keeps Indians of this vicinity in Such Continual alarm, that they Can[n]ot turn their attention to any thing but the care of their family and that they do but poorely. While the powerful tribes from Vancouver’s Island harass them in this manner, little hunts Can be expected from them and unless the Company Supports them against those lawless villains little exertions Can be expected from them.––James McMillan (McMillan and McDonald 1998:57)

These journals are from the first years of the Hudson Bay Company’s (HBC) Fort

Langley, established in 1827 on the south bank of the Fraser River about 45 km upstream

from the mouth (Maclachlan 1998). Prior accounts were valuable but they were

invariably based on brief encounters and the explorers and traders knew and reported

little of the cultural context of their interactions. In contrast, the daily logs at Fort

–– 73 ––

Langley, written by the Company’s proctors represent the first recordings of everyday

life, allowing us to see unfolding patterns of interaction from 1827 to 1830. From Fort

Langley’s vantage point, the Fraser River was a transportation thoroughfare, a Coast

Salish Main Street, with the fort as a major attraction and transportation and trading

node for the Coast Salish canoe travelers. People regularly came from Vancouver Island,

the upper Fraser, Puget Sound, and elsewhere; many to trade furs (prizing beaver),

sturgeon, and salmon at the fort, but more so to reach seasonal camps or to visit for

feasts or other gatherings. Men at the fort married local women, which they viewed as

advantageous “alliance[s]” (Maclachlan 1998 [e.g., June 23, 1829]). As Wayne Suttles

(1998) noted, the journals are “especially valuable,” for their descriptions of conflict,

raiding, and warfare.

Carlson (2001), in an analysis of the journals found that there were notations of at

least thirty conflicts during those three years. Fights erupted even at the Fort’s gates and

they noted constant alarms about the presence of the feared Lekwiltok, the southern

Kwakwaka'wakw who often raided southward into Coast Salish areas. The nature and

extent of these conflicts range from personal fights escalating into tribal ones, raiding

parties to punish individuals for “bad medicine” (April 26, 1828 [McMillan and

McDonald 1998:60]), to the scattering of women and children to hide in the woods upon

a raiders’ approach, and reports of gathering tribes to conduct retaliatory attacks on

enemies such as the Lekwiltok (e.g., around “500 men”; September 21, 1830 [McDonald

1998:159-160]). Undoubtedly, contact had already altered the nature of these groups

before the Fort was built, and later the Hudson’s Bay Company men at Fort Langley

hoped to change aboriginal culture even further to their economic advantage. The

traders even intended to minimize warfare amongst groups to allow for less

obstructions to hunting pursuits. They were more willing to sell firearms for use in

hunting than for warfare (e.g., McDonald 1998 [1829]:111-112; Angelbeck 2007:271-272).

The journal accounts such as James McMillan’s entries from 1828 provide the earliest in-

depth descriptions of warfare in the region:

–– 74 ––

Thursday 13th [March]. This morning a war party of Cawaithens Headed by Lammus passed up. They Say they are going to kill the Chiliquiyouks a tribe that lives on a Small river that Come[s] in from Mount Baker. The man that Stood watch last night observed a large Canoe full of Indians Coming on Slily till they were opposite to the Bastion but perceiving they were discovered they about Ship at once. We Suppose it was those vagabonds on the look out if every thing was quiet in the Fort, and take us for Chilqueyoukes. They are 150 men in ten Canoes, and ugly looking Devils they are––painted to their very ears.... A little after the war party left this they met Shientin the Musqueam Chief with his wife and two of his daughters. The war Chief took the eldest from him, menacing if he did not keep very quiet he would kill him & make Slaves of his family––two very fine looking girls.––James McMillan (1998 [1828]:56)

Accounts from the 1840s and 1850s

After the establishment of Fort Langley, there were several other expeditions

with extant accounts that have provided glimpses of Coast Salish warfare. Charles

Pickering (1854:15-16), on a global expedition aboard the Vincesses, described his visit to

a stockade near Dungeness Spit on the Olympic Peninsula, Klallam territory in 1841:

[In] the vicinity of Discovery Harbour, I was fortunate enough to fall in with one of the permanent stockaded villages. It was built in a concealed situation, on the bank of a small stream of fresh water, that afforded access by canoe; and it was not far from the anchorage at Dungeness. It appeared to be the proper home of all natives we had seen within many miles; amounting, perhaps, to as many as three hundred persons.In one of the houses I witnessed the remarkable treatment to which the Chinook25 infants are subjected; being confined to a wooden receptacle, with a pad tightly bandaged over the forehead and eyes, so that it is alike impossible for them to see or to move....Some of the men had their faces blackened, and I thought at first they were not pleased with my visit. However, I was conducted freely about the villages....On returning to the ship, I observed a skull lying on the beach; a circumstance that surprised me, as I was aware that these tribes take much pains in the disposal of their dead. On pointing it out to my attendant native, he looked sorrowful, and made some gestures which I thought referred to the common lot of mortality. He also showed me the marks of a wound, received by him, as well as I could make out, in an engagement with a Northern tribe (Pickering 1854:15-16).

25. Having just come from the Columbia River area, Pickering appears to have misapplied Chinook as having broader scope.

–– 75 ––

A few years later, in 1847, Paul Kane, journalist and artist, visited a fortified

Klallam village near Port Townsend called I-eh-nus. This was not long after the fort had

suffered an attack by the Makah. He described the fort as a:

... [D]ouble row of strong pickets, the outer ones about twenty feet high, and the inner row about five feet, enclosing a space of 150 feet square. The whole of this inner space is roofed in, and divided into small compartments, or pens, for the use of each separate family (Kane 1971:251).

While going northward through Puget Sound, Kane also remarked on

several defensive locales. Kane and his crew were even fired upon from a

stockade with “two stout bastions of logs” on the west side of Whidbey Island

(Kane 1859 [1847]:227).

William Ebrington Gordon, while aboard the HMS Virago, recorded a visit to a

fort in the southern Gulf Islands in 1853:

In passing the [Cowitchin] Gap [Porlier Pass] there is an Indian village prettily situated on the right hand shore with potatoe grounds sloping to the water on either side, where through the Gap the Cowitchin war village may be seen on a low point about 4 miles to the northward [Shingle Point]. It is a stockade built in imitation of Fort Victoria....

The next day we had to contend against a strong breeze and continuing tide, so we shored at the Cowitchin war village, and did not arrive at the Gap till about 3.30 PM. The war village ... at the time of our visit was deserted (Gordon 1853).

It is likely that this was the fort visited by Bishop Demers in the early 1850s. A

missionary history described an account, based on his notes of the encounter, which

appears to be the location of a fortification on Shingle Point, Valdes Island:

The Bishop saw here for the first time an Indian fort. This one which enclosed all the village cabins, was built as a protection against the incursions of the terrible Yougletas from the lower Fraser. The Bishop describes it as being about two hundred feet by fifty and surrounded by posts twenty feet high. At regular distances, enormous tree trunks were sunk deep in the ground to solidify the encircling posts which were much shorter. On top of the tree trunks there were figures, supposedly human, but in them would be difficult to say whether the grotesque or the ridiculous prevailed. “The best thing for you to do,” says the Bishop in describing them, “if you wish for a better idea than I can convey, is to blend both adjectives. They were ridiculous and grotesque” (Theodore 1939:187).

–– 76 ––

That fort was taken down by Chief Joe, at the request of Bishop Demers––he

wanted to encourage peace with the Lekwiltok, although in this example, the

“Yougletas”26 are misplaced as hailing from the Lower Fraser.

The Memoirs of Samuel Hancock

Initially setting out from Missouri in the 1850s, Samuel Hancock (1860, 1927)

prospected for gold in California, but soon turned his attention northwards to seek his

fortune in coal in the central Northwest Coast. While most colonists remained close to

the colonial forts, Samuel Hancock set out often on his own, hiring natives from local

groups as guides and travelling throughout Puget Sound, the Strait of Juan de Fuca, and

the west coast of Washington and Vancouver Island in the early 1850s. Because he

operated on his own, without an entourage of other colonists, expedition or military

crew as most previous encounters had been, he was able to experience and describe

aspects of Coast Salish, as well as Makah and Nuu-chah-nulth cultures, in a more

personal way. Hancock encountered these groups on their terms, not on the deck of a

ship, from the shelter of a trading fort, nor encumbered with a large entourage of other

colonists or traders. Instead, he engaged them as an individual and stayed as a guest in

their villages and homes. Other individuals also must have operated independently

among these cultures, but Hancock provided an account with his detailed memoir of his

travels and explorations along the Northwest Coast.27

In the early 1850s, when he first set out from Fort Seattle, he hired a crew of

seven natives, likely Duwamish, as guides. He handed each of them muskets and

26. There are numerous names used for the Southern Kwakwaka’wakw. I use the spelling “Lekwiltok,” after Duff's (n.d.) manuscript of the group which covers the time period of their conflicts with the Coast Salish. Other common names include Euclataws (Duff n.d.), Ne-cul-ta (Kane 1859 [1847]), yǝkwiłtax (Elmendorf 1993), Yukletas, Ucultas, among others (Hodge 1913). Today, descendant communities use the spelling of Laich-kwil-tach.

27. Hancock’s memoirs (1860, 1927) are rich in ethnohistoric detail, much like Jewitt’s account of his slavery to Maquinna in 1803. In fact, Hancock also was held prisoner by a Nuu-chah-nulth group for many days until his Makah friends interceded on his behalf and he was freed.

–– 77 ––

described their reaction:

The first night I encamped at the mouth of the Puyallup River where I discovered my Indians were very proud of their muskets with bayonets; it was quite amusing to see their manoevers with these arms, of which they had never seen any before, and they told me they intended having a fight before we returned, and capturing some slaves (Hancock 1860:95).

Not long after that encampment, Hancock and his men, canoing alongside

Whidbey Island, witnessed what appeared to be a battle at a Snoqualmie village:

That day we came in sight of a large party of natives who seemed to be fighting; I enquired of my Indians [guides] what it meant or whether they were really fighting? They replied that they were “Hias Silex” that is very angry. We ventured near enough to be able to see what they were doing without seeming to attract their attention; they continued charging and rushing through the midst of the crowd; apparently fighting, scalping and killing each other; I could see them strike with knives when one would fall as though deadk [sic], when his antagonist would spring upon him knife in hand and go through all the formula of scalping him at the same time cutting off a bunch of hair which he would hold up exultingly representing the scalp of an enemy; then rushing at another in the most furious way, he would perhaps stumble and fall when his adversary would serve him in like manner. This was a mock fight often up for the purpose of preparing themselves for a moment against an enemy and when I understood there was nothing serious in all this I enjoyed it very much. Soon they finished this pantomime and beckoned us to come to shore (Hancock 1860:97-98).

Once at the village, Hancock (1860:98) learned that the mock battle was

conducted to ready for a battle to take place the next day. A chief in the village was

preparing to attack the Snohomish. They had captured one of his wives, “leaving the

poor fellow only two wives in hand.” As Hancock and his crew of seven had muskets,

the chief encouraged them to join their party in the morning attack. Hancock did not

want to attack the Snohomish and declined to lend his muskets to them as well. He did,

however, offer to come along the next day to help broker a peace. Hancock described

the war preparations that night:

The old chief then blackened his face, and kneeling down made a most lamentable noise, something like a forced cry, at the same time making all sorts of gyrations with his hands; during this time the rest of them were painting themselves in a way I supposed for war, but as they pledged me their word to a compliance with their own proposition, I did not attach much importance to this, after finishing this painting they all got their clubs, knives and implements of war, and formed themselves in a circle around the chief, jumping, yelling and dancing most furiously for about a half an hour, the chief seeming to participate fully in the

–– 78 ––

excitement. When they ceased I asked what it meant, they replied “Cultus” (Hancock 1860:100).

The next morning, the war party of about a hundred men advanced on the

Snohomish village where the chief’s wife was held, encountering many warriors armed

with bows and some guns, ready in the bushes. The Snoqualmie Chief brought tobacco

and pipes as a offering for the return of his wife, which the Snohomish Chief appeared

willing to accept, perhaps because of the size of the Snoqualmie war party. Ultimately,

after much discussion, the Snohomish Chief accepted two blankets and two muskets for

her return. Then, they “joined each other seemingly forgetful that they had ever been in

opposition” (Hancock 1860:102).

From Hancock’s accounts, it is clear that villages were often friendly, especially

when members of his hired crew would know or be related to members of those they

encountered or visited. However, it is also certain that tensions abounded throughout

Puget Sound and beyond. His guides were always wary of other canoes while on the

water and suspicious of people they encountered on trails as they travelled throughout

the region. Hancock’s native guides would warn of unfriendly groups, as when they

were on the prairies near the trails that the Yakima used, near Snoqualmie Pass

(1860:125-128). Upon hearing this, Hancock became upset that they had selected this

spot to camp. They decided to move their camp to a spot less open: “bundling up our

effects [they] led the way some distances off the trail, when I spread my blankets and

slept soundly the Indians watching by turns all night” (Hancock 1927:128).

A Scalar Approach to Accounts of Warfare

Expressions of Increasing Modes of Power

From the above survey of historic accounts, it is apparent that these contain a

lode of information germane to this inquiry into Coast Salish warfare. This overview

has not meant to be exhaustive by any means, but merely intending to highlight

pertinent accounts from the historic records. Wolf’s (1990) conceptualization of modes

of power can provide a framework to array the types of conflict documented. Wolf

–– 79 ––

meant power in both physical and non-physical ways––one has the power to kill or to

remove from a position, as from a job in an institution, or from a role in a volunteer

organization; however, as we are are looking at conflict, the physical applications of

power typically will dominate.

(i) Power as an attribute of a person, or individual power

This form of power is intrinsic in the individual and does not involve interaction

with others on its surface, although it can manifest in relation to others, particularly in

how an individual displays power to others. Nearly all of the accounts record some

form of warlike display, symbolism, or disposition, all of which is meant to display

spirit power, the power that allows a person to be a great warrior. This type of power is

what led the Spanish to describe the Coast Salish in the Juan de Fuca Strait as “warlike.”

Even when they did not witness war directly, personal displays of power demonstrated

a war-like propensity. Oral histories and ethnographies, document an important

connection between a warrior’s success and the vitality of his spirit power; a connection

that will be discussed in more detail later. To an early explorer or settler, one likely

would not have needed to know about the cultural significance of spirit power to get the

message that it was also an expression of an individual’s physical power. Samuel

Hancock witnessed such displays of individual power at the warrior dance on Whidbey

Island, just as Simon Fraser had seen among the Musqueam. In such accounts, displays

of one’s power (i) are intended to show that one has further powers to dominate, injure,

or kill another (ii), as follows.

(ii) The ability of one to impose its will on another, or relational orinterpersonal power

This type of power is visible primarily in relations between warriors and their

captives, or chiefs and their slaves. While chiefs and elites were not able to force

members of their households to do things, they could impose their will on their slaves.28

28. Major works on slaves in the Northwest Coast include Mitchell (1984), Donald (1997), Ames –– 80 ––

As Ames (2001:1) described, “Slaveholders not only controlled the labour of slaves, but

had the power of life and death over them as well.”

Slaveholders’ power over slaves typically originated from their initial capture in

warfare, although they could be traded or purchased thereafter. In itself, the act of

taking a slave, is the act of one enacting power over another, with the end result of

control, wounding, or killing. All historic accounts of warfare and conflict involve, by

its very nature, the attempts for the imposition of one’s power over another. Even at its

smallest scale, such as with a fight between two individuals, the goal is the imposition of

power. The following account of a “row” outside Fort Langley illustrates this point:

Sunday 10th [January 1830]. Another row amongst the Indians of our neighbourhood––this afternoon one of two Quaitlines that Came down upon a Special visit to the Musquam Village right opposite to us, was brought to our wharf lifeless with 7 or 8 arrows Still Stuck in his body & otherwise much mangled with the Knife––this butchering now is in revenge for the death of the old Musquam that was killed by his Son-in-law in the upper Village latter end of Novr.––nor is the difference likely to end here––One of our men with two of the women happened to be in that direction at the time making ashes: The poor wretch on being mortally wounded made an effort to throw himself into their Canoe, but his pursuers were too much bent upon their purpose to be defeated by this Screen. One of our people was present during the affray (McDonald 1998:136).

Events involving just two people or a few can escalate into yet larger scales, or

into higher orders of the application of power, as in the case of the above-mentioned

event.

(iii) The control of social settings, or organizational power

Monday 11th [January 1830]. A good deal of reconnoitering going on all night between the two villages––In Course of the forenoon the Quaitlines amounting to about 60 men in 12 Canoes came down armed best they Could, and Seemed to muster from 10 to 12 Guns of one Sort or another––They made it a point to Call upon us first & tendered us a large Sturgeon for ammunition, which we refused them for a variety of reasons.... They then wished one of the Gentlm. to accompany them to the Musquam Camp––this was also refused from the Same motive.... Tuesday 12th. The trouble Continued on the other Side till late today––

(2001), Ruby (1993), and Todd-Bresnick (1984).

–– 81 ––

Ever since the Quaitlins Crossed they fired occasional volies [volleys], which, with other Signs of hostility did not indicate a peaceable disposition––they now tell us that two of the Musquams are Slain––which we doubt much, as with them I find a man is dead when he acknowledges his life in the hands of his enemy! When the Quaitlines Came down yesterday they told us they had already killed 2 Musquums above, which on further inquiry proved to be a death of this mild nature.29

Wednesday 13th. It is ascertained that two Indians were actually killed & one of them very luckily the identical man that took the life of the other on Sunday––here ends the business for the present.––James McDonald (1998:136-137)

In these entries, a murder by one spiraled into an event that involved scores of

Kwantlen men and the whole village of Musqueam on the other. Initially, a murder of a

Musqueam man by a Kwantlen, followed by the Musqueam killing a Kwantlen man,

escalated to a battle involving 60 men and two further deaths, including “luckily” one of

the prior assailants. Here, with this escalation, we have the display and execution of a

broader mode of power. The Kwantlens, overnight, organized dozens of men to attack

the Musqueam.

As Samuel Hancock (1927 [1860]) had witnessed on Whidbey Island, a

Snoqualmie Chief was able to assemble 150 men quickly in order to attack a Snohomish

village. The presence of the multitude of war canoes was enough to encourage the

Snohomish Chief, who held the wife captive, to negotiate. With the Kwantlen vs.

Musqueam account, it is interesting to note that even with 60 men on one side and

presumably as many on the other, the battle occurred at a distance, with the majorities of

each group on the either side of the Fraser River––and only two individuals died. That

is, much of what occurred during those battles concerned the chief’s display of the

29. From McDonald's (1998 [1830]) account, it is interesting to note from his experience that when one “acknowledges” one is in control of another in battle––that is, submitted to their interpersonal power (ii)––it is regarded, spoken of, as death. This relates to how when a non-elite is captured, it is often considered a social death. Elmendorf (1960:346-347) has stressd the “immutable slave status” imparted a “a slur against their status comparable to ‘slave blood’ in an upper-class kin line.” This often did not occur for elites as they were valued for their price in ransom, whereas a non-elite only had value as a slave commodity; moreover, elites could hold a feast to clear their name, display and reaffirm their status (Elmendorf 1960:347).

–– 82 ––

organizational power (iii).

Despite a substantial amassing of warriors for the cause, the main impetus was

to show that one has power and can implement it––and they might kill one or two

people to show that their organizational power is not a bluff. However, at least among

those Coast Salish groups, it appears that they did not want the tension to escalate.

These accounts suggest that there is a lot at stake if conflict develops further, taxing

alliances, intermarriage ties, and disturbing the peace. For such reasons, perhaps many

conflicts at this scale involve lots of bluffing displays instead of the actual

implementation of that power or violence, a situation that may indicate that bluffing was

less used or would have been less effective in interactions with non-Salish groups, with

whom such shared interests or ties were much less common.

(iv) The ability to establish or demolish the settings themselves, or structuralpower.

The next dimension of power demonstrates the ability to organize within a

sociopolitical field and to alter the conditions of that field itself. Warfare at this scale

might result in the destruction of a village, or its resettlement to a more defensive

location––in effect, altering the fundamental nature of how the social setting is

constituted. Rozen’s (1985:108) Hul’qumi’num informants reported that “the so-called

‘wars’ with the Southern Kwakiutl only took place on a large scale (i.e. large enough to

force abandonement [sic] of exposed villages on the Gulf Islands) from about 1790 to

1850.” Similarly, Cook (1979) described how the Chilliwack people had moved from the

Chilliwack River Valley to the Chilliwack Area after the establishment of Fort Langley,

which she argued was partly done because the traders afforded them protection from

the “more warlike Cowichans and Qwantlens.”

Warfare not only led to resettlement, but also contributed to substantial

destruction of existing villages, including both structures and inhabitants. For instance,

Duff (1952) learned from one informant that the Nlaka’pamux (formerly Thompson)

raided the Chilliwack and burned a long plankhouse––the village was named yukyuke'us

–– 83 ––

(Yakweakwioose), or “burnt out.” Furthermore, Duff (1952:96) recorded an account in

which a Sumas warrior, KwEl, led a retaliatory raid on the Lekwiltok, breaking into their

fortified house and killing all the warriors there.

In 1856, the first European expedition to attempt to cross Vancouver Island to the

West Coast, led by Adam Horne, scouted a Haida war party coming down from the

north as they were approaching the mouth of the Qualicum River. They hid until the

party passed southward, without noticing them. I present this at some length as it

provides an eye-witness account of such a devastating attack on a Qualicum village:30

It must have been six o'clock next morning, or a little later, when the Iroquois31 aroused me, and told me in a subdued voice, that we were within one mile of the Qualicum, and that for some time, he had been watching a large fleet of northern canoes approaching the creek. What they intended doing, of course, he did not know, but he anticipated trouble....We waited patiently to see whether those Indians would return or not. It was fully twelve o’clock before the first of them came into view in the lower reaches of the creek. We were horrified at the antics of these demons in human shape, as they rent the air with their shouts and yells. One or two of those manning each canoe would be standing upright going through strange motions and holding a human head by the hair in either or both hands. The wind at this time was almost blowing a hurricane from the north, and the sea was tipped with angry white caps in every direction. Turning the prows of their canoes to the south, these northern Indians hoisted mats as sails, and fairly flew along before the gale. In an hour’s time they were all out of sight behind a bend in the shore line.... After lying concealed another hour we once more launched our canoe, loaded it up with our supplies and impedimenta, and poled our way along the shallow beach towards what we were now convinced was the mouth of the Qualicum.... In case we met with any natives, who might give us a hostile reception, all of our men had their muskets loaded and lying by the sides. We saw nothing of the rancherie on entering, but

30. This account is by W. Wymond Walkem (1914), who provided quotations for many passages, but oddly remained in first person (in this case, Adam Horne) outside of quotations; He is not clear on its sourcing. The boundary between Adam Horne’s words and his own is also unclear. Despite these qualifications, the expedition is well-documented (e.g., Hayman 1989:32, 48; Akrigg and Akrigg 1997:116); the lake just upriver on the Qualicum is named after Horne. But, more importantly, the descriptions accord with what is known about Northwest Coast warfare.

31. This is a one-armed man, part Iroquois and Chinook named Tomo, who was a participant in several expeditions on the island (Hayman 1989:32). Robert Brown, for his expedition years later, also employed him, noting that “One armed Tomo (his name is ‘Toma Antoine’ or ‘Thomas Anthony’) [is] of mixed Iroquois & Chinook origin but undistinguishable from a half-breed, & noted as a linguist & hunter....” (Hayman 1989:48).

–– 84 ––

volumes of smoke were still pouring out from one side of the stream beyond a projecting point, covered with heavy timber. In five minutes we were round this point, and then a most desolate and pitiable condition of things met our view. What had evidently been a rancherie was now a blackened heap of burning timbers. Naked bodies could be seen here and there, but not a living being was in sight. Our interpreter called out several times that if there was any person living to come out––that we were friends, and would do them no harm. He got no answer, except the echoes from the surrounding hills, and he then walked over to where the lifeless bodies were lying. Horror of horrors! Every trunk was headless and fearfully mutilated. We searched the surrounding underbrush for living beings, but without success. Discouraged, we sat down upon a drift log to discuss what we should do. Some of my men were returning at once to Fort Victoria, but this I positively refused to do.... There were no Qualicum Indians from whom I could gain my information, so I must try and find the trail without assistance. If there were any left they must be prisoners in the hands of these northern Indians.

They eventually found a woman, badly injured, who related the attack to Tomo, the interpreter:

They had all been asleep in the large rancherie when the Haidahs crept in with stealthy step, and more than half of those asleep were killed without awakening. The remainder were quickly killed, there being five Haidahs to one of themselves. She was wounded with a spear, but had seized a bow and fled to the side of the creek and had hidden herself beneath the bank. The Haidahs had taken away with them two young women, four little girls, and two small boys. This expedition was in revenge for the killing of one of the Haidahs when attempting to carry off the daughter of one of the principal men who live where the death currents meet (Cape Mudge). Beyond this we could get no more information.... This camp, with its headless bodies, was no place for us, so we returned to our canoes and left the creek as we had entered (Walkem (1914:41-42).

But, not all such enactments of structural power were restricted to

intercommunity battles with non-Coast Salish peoples. For instance, one of Jenness’

(1934) informants described how the Comox similarly burned all three houses of a

Saanich village. And, McMillan (McMillan and McDonald 1998:57) in 1828 described a

devastating attack by Cowichans on upriver groups like the Pilalt, quoted above (see

page 73). In these accounts, it is clear that such attacks altered the social setting of the

victims, impeding their ability to engage in routine economic activities, often to the

dismay of the traders, who relied heavily on the local production of food and trade

goods such as furs.

–– 85 ––

More examples of such destructive power are illustrated in the accounts of the

wars with the Lekwiltok. In many of these accounts, the attacks were similar to those

that took place between Coast Salish communities. For example, on August 11, 1827,

some Skagit people came to trade at Fort Langley, partly to provision for their trip

northward to rescue or punish the Lekwiltok for the capturing of two of their women

(Barnston 1998:31-32). On May 8, 1828, the Lekwiltok attacked a Cowichan summer

village and killed a Musqueam chief. Other Lekwiltok attacks were more destructive: on

June 12, 1828 (McMillan and McDonald 1998:65), they attacked a Musqueam village,

reportedly killing 3 men and taking or killing 30 women and children, completely

altering the nature and composition of the village; and on July 21, 1830, the Lekwiltok

attacked a village at Point Roberts, wounding four Cowichans and killing a

Snuneymuxw (formerly Nanaimo, Snanaimuq).

Taylor and Duff (1956) documented that the Lekwiltok took over territory in the

Northern Gulf, overtaking control of lands as far south as Quadra Island’s southern tip,

Cape Mudge. They drove the Comox from their villages and seasonal camps. Boas

(1889) documented a Snuneymuxw and Sechelt attack upon a Lekwiltok village as far

north as Salmon Bay on Vancouver Island presumably after the Lekwiltok had taken it

from the Comox.

There are other accounts of the movement of entire villages. Jenness

(1934) described how the Songhees (also Songish) and Saanich moved their

settlements because of the regularity and intensity of raids:

It was through fear of both the Comox and the Kwakiutl that the Songish retreated in summer above the gorge at Victoria, and the Saanich sent their women and children to secluded spots during May and June, the usual seasons for raids, while the men maintained a nightly watch on housetops. During the 19th century, indeed, the Saanich abandoned one of their villages near Sidney, on the east side of the peninsula, and moved to Patricia Bay, on the west side where they were less exposed to attack (Jenness 1934).

Similarly, the Klahoose, according to their Chief Julian, maintained village sites

deep up Toba River that were less accessible rather than be open to attack in Toba Inlet

(Black, Urbanczyk, and Weinstein 2000:32). Many Coast Salish oral histories detail –– 86 ––

significant alteration of lifeways resulting from Lekwiltok attacks.

Once the [Cowichan] were at war with the tribes on the American side of the Straits. While they were absent from their villages some of the Kwakiutl bands swooped down upon their settlements, burnt their houses and carried off the women and children into slavery. When the [Cowichan] warriors came back they found their homes destroyed and their families carried off into slavery. Nothing was left to them but the smoking remnants of their dwellings. Not even a dog remained (Hill-Tout 1978:160).

After that event, the Cowichan decided to organize the Coast Salish together in

order to put an end to this cycle of warfare. They held a council of war, inviting chiefs

from numerous Coast Salish groups from the Nanaimo and Sechelt in the north, Sooke

to the west, Klallam, Skagit, and Nisqually and other Puget Sound groups, as well as

some from the Fraser River. Setting scouts along the Strait of Georgia, they were going

to be prepared for the next time the Lekwiltok returned. Finally, a battle occurred at

Maple Bay, between Vancouver Island and Saltspring Island. From the numerous

accounts that are available, most if not all the Lekwiltok warriors were killed; I describe

the battle in more detail in Chapter VIII.

Although the oral histories may have exaggerated some of the specific outcomes,

they are consistent in noting that these battles ended the Lekwiltok raids. The very

large-scale organization of Coast Salish groups for the purpose of ending the Lekwiltok

raids is a clear example of structural power that permanently altered the nature of

Lekwiltok/Coast Salish interaction. For decades prior, evidently since the 1790s

through at least the 1830s––the years of the Lekwiltok wars––it could be argued that the

environment of constant raiding and counter-raiding was the predominant sociopolitical

field. Most attacks during that period, while destructive, could be described as

examples of organizational power, whereas, the Battle at Maple Bay, altered the

sociopolitical field itself.

A final example of structural power is in the actual extermination of a people. In

the Gulf of Georgia, this happened to the Chemakum, said to be a “troublesome” group

inhabiting Port Townsend. According to Gibbs (1877:191), the Makah first attacked

them fiercely, after which the Chemakum had to battle with the Snohomish. Finally, –– 87 ––

Chief Seattle led the Suquamish against them, destroying their fort and nearly killing

them all. The Chemakum survivors joined surrounding groups like the Klallam (Curtis

1970 [1913]:142; also Elmendorf 1993:143-145).

These accounts demonstrate that the Coast Salish did not always directly enact

the next scale or mode of power; rather, they frequently only signaled their ability to do

so. In effect, they were threatening (or bluffing) that the next higher level of engagement

would be undertaken if the enemy did not back down. One can be “warlike,” through a

show of force and readiness, putting on various “grotesque” faces and engaging in

menacing acts, signaling both power and willingness to conduct war. Alternately, as

Simon Fraser had seen, one could bear a “belt of scalps.” Such displays demonstrated

that warriors had successfully killed others (ii) many times before. Slaves also were

symbols of status. That status resulted from a clear demonstration of control of another.

Similarly, other dimensions of power will also have their subsequent material residue in

some form of symbolism.32

A “Continual State of Fear”: The Power of the Lekwiltok

Based on this frame involving the different modes of power, it appears from

historic accounts that the Lekwiltok were more powerful than the Coast Salish. They

were formidable and considered a menace to all groups they encountered. They even

called themselves by names that could instill fear, as one group’s name meant

“unkillable thing,” after a worm that would keep squirming even when split into pieces;

other group names included “murderers” or “the angry ones” (Curtis 1970

32. One can view elite traded items, described at potlatches, as symbols of those relationships, symbols for alliances that one can call upon––that is, organizational power that one has access to. Often, these items are discussed by archaeologists as ceremonial or otherwise non-utilitarian, but these items serve a potent function: to demonstrate the social capital that one can draw upon when necessary––a degree of power that perhaps could be used to alter the social settings itself.

–– 88 ––

[1915]:308-309). Crosby (1907:68), an early missionary, described how the Lekwiltok

attacked all who travelled through the narrows between Quadra Island and Vancouver

Island, or what he referred to as the “gauntlet”:

The northerners were not always successful in making the trip home with their booty. The Cowichans would gather at Dodd’s Narrows and Active Pass, or at Cowichan Gap, and set upon the victors, often turning their victory into defeat. If they escaped the Cowichans they still had to run the gauntlet of the Yu-kwul-toes [Lekwiltok], the most to be dreaded of the whole coast tribes, and many a Tsimpshean, Hydah or Kling-get war party has found its death trap at Seymour Narrows or the Yu-kwul-toe Rapids.

In the Fort Langley journals, there are numerous references to the Lekwiltok and

the general fear of them. As McMillan (McMillan and McDonald 1998 [1828:65) noted in

his journal entry, after the Lekwiltok had attacked the Musqueam, “The Country

her[e]abouts is in [a] Continual State of fear by their powerful and Blood thirsty enemies

from the Gulf of Georgia and Johnston’s Straits.” Similarly, McDonald (1998:101), on

Friday the 13th, March 1829, remarked that “It is impossible to describe their Continual

alarm at the very name of this formidable foe.” Yet, as I have discussed in more detail

elsewhere (Angelbeck 2007), this fear of the Lekwiltok appears to be justified, for several

reasons, each of which involves a Lekwiltok advantage in dimensions of power,

primarily in organizational and structural power.

Kwakwaka’wakw peoples such as the Lekwiltok had warrior sodalities, or secret

societies, which Mitchell (1989:5-6) stated was a “superior organization for fighting,”

and which was distinctive from Coast Salish modes of social organization, which he

described as “atomistic.”33

The relative autonomy of extended households in all economic and political matters has contributed to the impression many ethnographers convey of the “atomism” of Salish society. Kwakiutl extended households, however, were additionally associated as members of well-

33. While Mitchell summarized the ethnography of the Coast Salish as conveying a sense of atomism, a term that characterizes the Coast Salish sense of individualism and autonomy. However, the term also connotes the isolation of disparate parts; therefore, the term anarchism, I believe, is more appropriate, as it conveys both the autonomous quality of households and individuals while having principles of organization that provide networks to fulfill certain needs.

–– 89 ––

defined descent groups––numayms––as important sub-units of the local group. These divisions were linked through the formal ranking system to provide a structure for Kwakiutl society lacking for the Salish (Mitchell 1989:6).

Further, he described how Salish winter dancers also displayed violent behavior,

however, these were the “performances of individuals––not members of a sodality....

among the Salish, the warrior was a feared, almost uncontrollable, and decidedly

solitary figure––fully in keeping with the mooted Salish atomism” (Mitchell 1989:9).

Here, Mitchell makes a strong point. As he noted, Boas (1897:664) himself attributed the

origin of secret societies to warfare, noting that those societies were very active during

periods of warfare and that the initiator of the ceremonies was a spirit named

Winalagilis, “the one who makes war upon the whole world.” While there may have

been clear war symbolism associated with these societies, it is not simply a matter of

ideological orientation to warfare they may have had, rather, the organizational power

was greater. This dimension of organization outside of the household consisted of an

institution ready-made for warfare implementation. Contrast that with the Coast Salish

practice where a warrior has to cajole others into joining his cause, perhaps by

promising participants a portion of the booty. It must have been easier to persuade

people to participate in the defense of their own village rather than to join an attack

upon some distant community, even if for revenge.

According to Roscoe (1993), practice theory accounts for how a powerful

individual can implement practices that eventually lead to an institutionalization of his

power. Leaders in many societies used numerous practices to directly and continually

create and maintain alliances, all of which are costly––some leaders ending up owing

their supporters more than they have credit, resulting in an unstable form of power.

Instead, it is better for leaders, when possible, to “institutionalize their dominance”

beyond their own charisma or ability to convince others. The secret society is an

example of such a practice, as it does not rely upon association with a warrior or an

alliance with a chief that may enable some benefit in return. Rather, the secret society is

organized under an ideology and set of ritual practices that orients the relationship of –– 90 ––

warriors. In the Lekwiltok case, the warriors form part of a ranked institution that

extended beyond and existed separate from the organization of households. Consider

the difference in effectiveness with organizational power: one can put one’s energy into

canvassing individuals to your cause in warfare, appealing to their self-interests or

convincing them of its justification––in such cases, the power to join is heavily in the

control of the free individual. Or, the secret society of warriors pursuing this attack can

demand that its members participate in the attack, or be barred from membership in that

institution––the weight is on side of the institution, or the secret society. For the

Lekwiltok, their warrior societies were already an institution––already organized and at

the ready––whereas Coast Salish war parties were formed after efforts of organizing for

each occasion.

In addition to greater effectiveness in mobilization of people, Mitchell (1989:5)

also recognized that their population was greater. This point is particularly relevant for

the period of the Lekwiltok wars, from 1790 to the 1830s, given that a smallpox epidemic

had decimated the Coast Salish beginning about 1782 (Harris 1994; Boyd 1999).

Smallpox affected populations at different times and in different intensities throughout

the continent (Dobyns 1966, 1983); and the Coast Salish were particularly hard hit

during the outbreak of 1782. In a map detailing the population from a census after the

smallpox epidemic (Figure 3), it is clear that their numbers were diminished relative to

other areas, particularly the Kwakwaka'wakw (Lekwiltok) to the north and

Nlaka’pamux (formerly Thompson) in the northeast. This suggests that the disease

spread with devastating effect within the Coast Salish interaction network and slowed

down in neighbouring regions. This is important here concerning the Lekwiltok, as

smallpox is not documented to have affected the Kwakwaka’wakw at the time. The

severity of their population decline undoubtedly affected Coast Salish organizational

power. This disparity in numbers and resulting social instability conferred an

advantage on the Lekwiltok in their conflicts with their southern neighbours.

The effect of technology must also be considered. Vancouver’s expedition in

–– 91 ––

! !

!

0 25 50

km2,000 1,000 500 250

BRITISH COLUMBIA, CANADA

WASHINGTON, US

Olympic Peninsula

Vancouver Island Strait of

Georgia

Approximate spread of smallpox, 1782

Population estimates from H.B.C. Censuses, 1830s

Figure 3: Map showing approximate spread of smallpox in 1782 and population estimates in the 1830s (Redrawn after Harris [1994, Figures 2 & 3]).

1792 documented how the Lekwiltok of Johnstone Strait were well armed with muskets,

while within the Coast Salish territory they had just crossed, they did not mention the

presence of firearms (Cole and Darling 1990:120-121). Prior to contact, all of the

Northwest Coast peoples fought with the same types of weapons: knives, slings, clubs,

spears, darts, bows and arrows––the same means of destruction were available to all

groups. After contact, firearms were introduced, however, it is clear that their

distribution was uneven and groups with access to the outer coast had greater access to

traders who could provide guns. Sea-based traders regularly coursed up and down the

outer coast, and they visited the inner passages such as the Gulf of Georgia much less

–– 92 ––

frequently.

The Lekwiltok, while not on the outer coast, regularly interacted with other

Kwakwaka’wakw groups that were. The Central Coast Salish did not have ready access

to fur traders until the establishment of Fort Langley in 1827, and that was a land-based

trading fort. Furthermore, prior to the establishment of Fort Langley, the fur traders

would not have been as interested in the types of furs the Coast Salish could produce

(e.g., beaver pelts), because during the early decades they particularly wanted otter

skins for the China trade. Sea otters were rare in the Gulf of Georgia region and this left

the Coast Salish communities at a disadvantage concerning trade goods (Kennedy and

Bouchard 1983:116; Suttles 1987d [1957]:155). Once Fort Langley was established, and

after the serious decline in sea otter populations throughout the Northwest Coast,

beaver pelts were the most prized furs, and these the Coast Salish could supply.

However, there was a major difference between the amounts paid by sea-based traders

and land-based ones at Fort Langley. As McDonald noted in 1829 (1998:111) when some

Cowichans wanted to trade furs for rounds of ammunition to attack the Lekwiltok, “the

natives of Vancouver's Island and all along the Coast Can have no difficulty in obtaining

elsewhere for their Skins ten times the quantity of amm. we give.” So, even once traders

were readily available to the Coast Salish, there was hesitancy to provide them with

firearms, partly in concern for their own welfare as they lived among Coast Salish

groups while sea-based traders moved on after exchanges (Angelbeck 2007:271-272).

European traders wanted to distribute firearms and ammunition to be used for hunting,

to bring in furs. These conditions of geography and historical circumstance led to great

disparities in access to firearms, resulting in power differences as well, as recounted in

oral histories. For example, Curtis (1970 [1913]:20) recorded two accounts of conflicts

between the Lekwiltok and Klallam:

One of the earliest wars of which the old men now tell began when a party of the Lekwiltok attacked a Clallum village on Whidbey [I]sland. The islanders, about to celebrate a wedding, were expecting the arrival of friends from other Clallum settlements, and seeing the canoes of the Lekwiltok coming ashore, they hurried down to meet their supposed visitors, never dreaming that an enemy would approach by day. The

–– 93 ––

northerners, armed with guns, of which the Clallum as yet had none, quickly opened fire, killing many and dispersing the rest in the woods, and then pillaged the houses.

He also documented that a retaliatory attack upon the Lekwiltok was also halted

by guns: “Again the allies assembled, and this time they found and attacked the

Lekwiltok. But the guns of the northerners were too much for them, and they turned

and fled” (Curtis 1970 [1913]:21).

The disparity in firearms access, in light of Wolf’s (1990) modes of power, can be

seen as altering the dynamic of warfare, which had been previously, with stone and

bone and shell weapons, a field of “equal footing” (Mitchell 1989). It can be seen as

giving them a power that altered the field within which warfare was conducted. In fact,

in reviewing these three primary advantages––rank institutionalized in secret societies,

population inequities after smallpox, and the disparities in firearm distributions and

access––two are the result of contact. That still leaves an advantage in readiness for

organizational power, between the institutional form of secret societies and more

autonomous form of the Coast Salish, described here as anarchic. However, that form of

organizational power in warrior societies could be matched, as Salish groups could

draw upon their networks of alliances when they needed. This is seen in the account of

the Battle at Maple Bay, where their organizational power (iii) superseded the Lekwiltok

and ultimately altered the setting for subsequent encounters, enacting structural power

(iv).

Colonist and Coast Salish Conflict

A common way to array the types of conflict, is to examine the increase in

conflict from murders and feuds to intra- and inter-community conflict. However, while

it is generally true that each of those conflicts is an escalation of the prior, those

categories do not encompass the complexity those classes, especially in regards to the

scale or degree of damage or the nature of power applied. This is most readily seen in

the conflicts between Coast Salish groups and traders or colonists. For the Coast Salish, –– 94 ––

most incidents involved rather small-scale conflict with the killing of one or a few

settlers. In the following missionary account, Reverend Ronden (1913) described the

response of colonists to murders, using a gunboat:

Time and again, whilst the painstaking missionary was variously engaged, the gunboat “Forward” was despatched from Victoria, now to put an end to some bloody affray between Indians of opposing tribes, now to capture and chastise some wild native guilty of slaughtering Whites. It is sad indeed to record that among the Whites that had commenced pouring into the country there were not a few who by their overbearing manners and dissolute conduct often provoked Father Rondeault’s parishioners to retaliate on them in a sanguinary way. However, it is gratifying to know that in these circumstances the Government wisely worked hand in hand with the clergy. Through Bishop Demers and Father Rondealt’s paternal influence, Cowichan was several times saved from being wiped out by bombardment. Upon the missionaries’ advice many a murderer spontaneously delivered himself to the secular arm, and resignedly paid the penalty, which generally was death (Ronden 1913:45; emphasis added).

First, this account reveals that missionaries were able to protect their First

Nations’ parishioners in some cases, but it implies that saving the villagers did not

always occur. Moreover, “bombardment” by gunboat was the response to a village for

the murder of a settler. One informant related to Thom (2005:255), how devastating

these gunboats attacks were:

It was the government that came along to this country and he told his whole crew “Take a pot shot. Lamalchi Bay.” Took a pot shot at Kuper Island Indian Reserve, took a pot shot at Leey'qsun tribes [Lyackson].... Took pot shots, killing the people. Many people died. Old people. Old women. Little newborn babies were slaughtered by the white man from these ships....You forgot that it’s your people that destroyed our villages. You shot at us, pot shots with those big guns, destroyed all the big houses. Indians had to go hide in the mountains from you people.

At Lamalchi Bay, the British reportedly killed six or seven Lamalcha and

wounded many more, however, they “completely destroy[ed] the buildings” (Victoria

Daily Chronicle 1863; cited in Arnett 1999:145). Thom (2005:256; emphasis added)

commented that the oral histories of the event revealed the “transformative power wielded

by the state in seizing Island Hul’qumi’num land”; he is describing the use of gunboats

as structural power (iv). It is interesting to note, in viewing Coast Salish/colonist

–– 95 ––

relations in this Wolfian frame of power, that colonists, when faced with a murder of one

or two settlers (ii) responded with an application of full structural power (iv). They

altered the social settings with “gunboat diplomacy” (Gough 1984), destroying the

village at Lamalchi Bay for instance, or coming in with an overwhelming show of force

with hundreds of militiamen, as at Cowichan Bay in 1856 (Gough 1984:63-67). Perhaps

it is an indication of their awareness of how outnumbered the settlers were prior to

widespread immigration. Perhaps they felt a need to always show their capability for

structural power.34 This structural power primarily resided in the technology of large

well-equipped gunboats, and European militaristic organization, which was even

stronger than the the numayum of the Kwakwaka’wakw, described by Mitchell (1989).

Earlier, for the HBC traders, a fear was present that the Coast Salish would

organize. Above, I mentioned McDonald’s (1998 [1829]:111) account regarding his

hesitancy to provide firearms or ammunition in large amounts to Salish groups that

lived near the Fort. As Roderick Finlayson (1879:68), Chief Trader of the HBC at Victoria

recorded, it was their intention to prevent such organization: The Policy of the company was honesty,––and also to keep the several tribes divided and at enmity among themselves. This plan was followed for purposes of protection to ourselves. ––In short to keep up a jealous feeling between the respective tribes (Finlayson 1879:68).

Conclusion

From the accounts, both historical and native, described above, several points are

evident regarding the effects of warfare in general and the nature of Coast Salish warfare

34. Similarly, to the far north, the Russian encroachment was met with resistance in the Aleut area. In 1745, Mikhail Nevodchikov lost 32 crew members at what is now known as Massacre Bay and Murder Point. In 1762, whole crews of ships commanded by Alexi Drujinin and Stephan Glotov were killed on Unalaska (Coppock 1970:iii). Two years later, Ivan Soloviev returned to avenge those deaths, and killed about 300 Aleuts. Coppock (1970:iii) further wrote that the intervillage alliances of the Aleuts were “ineffectual” as “the Russians pushed across the area village by village and island by island, they eliminated approximately 80 percent of the Aleut population.” Thus, in response to the Aleut attacks, the Russians responded overwhelmingly, nearly wiping out the Aleuts in the process.

–– 96 ––

in particular. First, warfare had substantial effects on indigenous cultures, particularly

with respect to what Wolf (1990) termed “structural applications of power.” Other

forms of power in warfare may have created fear, alteration of individual lives (through

enslavement or wounding), or the reorganization of households due to deaths of

individuals, however, warfare had structural effects that altered peoples’ ability to

engage in other cultural practices, including where and how they lived. On seasonal or

subsistence outings, conflict was part of peoples’ calculus for determining where they

decided to camp. For example, Louis Pelkey, one of Suttles’ (1949 [6]:65) East Saanich

informants, described how the “old people used to like to camp where fire not seen and

would pull canoe up into bush.”

Second, it affected how they conducted daily routines. For example, one of

Smith’s (1940:159) Puyallup-Nisqually informants stated: “My grandfather went out in

his canoe to get wood. He always had his arms near him.” Suttles (1949[5]:70) noted

Julius Charles’ statement that “Lummi didn’t go [for the] month [on a seasonal outing].

Had to build forts to protect the people.” That is, they opted not to conduct routine

annual subsistence activities, and lose any potential surplus stores, in order to build

these defensive sites. As Collins (1974:122), noted from her work among the Upper

Skagit, “the blood feud not only interfered with economic activities but also affected the

round of attendance at religious ceremonies and potlatches." Since warfare structurally

affects how and where their lives and cultural practices were carried out, it becomes an

important context to consider for any study, archaeological or anthropological,

concerning those periods when warfare was conducted.

It is also clear from these accounts that warfare, or at least its increase in

intensity, was a by-product of the contact encounter. Even before the Spanish and

British expeditions reached the Gulf of Georgia region in the early 1790s, contact had

substantially affected the Coast Salish through the spread of smallpox. The introduction

of firearms also tipped the balance, leading some like the Lekwiltok to play those events

to their full advantage. The introduction of the fur trade established new dynamics into

–– 97 ––

the economic systems of the Northwest Coast, a matter which we will return to in more

detail in Chapter X.

While these colonial period accounts of warfare are especially valuable, they

cannot be seen as a mirror of warfare prior to contact. As Ferguson and Whitehead

(1992) emphasized, each and every postcontact account of indigenous warfare is, in fact,

an account that is affected by the nature of contact. In a real sense, there are no accounts

of pristine, indigenous warfare unadulterated by contact with Westerners; this is similar

to a point repeatedly stated in the 1990s about the history and conditions of ethnography

in general, all of which were conducted in colonial situations. For this reason, Haas

(1990) argued that an archaeology of warfare is critical to an understanding of

precontact warfare.

Considering the effects of contact, through Wolf’s (1990) modes of power,

perhaps the anthropological discourse of the Coast Salish is inadequate; as Bierwert

(1999:15-18) has summarized, the literature often presents the Coast Salish as a “raided

people rather than raiding people.” This is a by-product also of reliance upon historic

accounts that highlight Lekwiltok superiority. Such narratives do not take into account

that there power imbalances, both organizational and structural, that resulted from

particular historical circumstances with limited duration. Moreover, the circumstances

of that early postcontact window exhibits points of contrast to the evidence for Coast

Salish traditions for warfare, a point returned to in the next chapters. Examples from the

historic accounts above provide evidence to the contrary of this notion of the Coast

Salish as a “raided people.” Accounts of the Battle at Maple Bay provide indications of

how, despite these postcontact disadvantages, Coast Salish groups were able to match

and supersede the organizational power of the Lekwiltok, destroying Lekwiltok villages

in the north, and transforming the nature of the Coast Salish-Lekwiltok cycle of violence.

–– 98 ––

Chapter V: Welfare and Warfare: Modes of Production and Destruction

While early historic records present a good deal of detail on warfare,

ethnographers have sometimes overlooked it. The traders at Fort Langley journals

could not avoid it (Maclachlan 1998). They were outnumbered and surrounded by the

Coast Salish and so fortified themselves. The Fort was threatened and occasionally

attacked (once, according to journals of James Murray Yale, by about 600 Lekwiltok

[Waite 1977:15-16; McKelvie 1957]). Coast Salish oral histories are replete with accounts

of warfare. Even their cosmological histories concerning the coming of or Xa’:ls (Khaals),

or as brothers the Xexá:ls, involved turning warriors into stone, which suggests that

warriors existed since the time of transformation (e.g., Jenness 1955:22). Among some

ethnographers in the past, however, this aspect has been underplayed, a point which has

been emphasized for the Northwest Coast as a whole (e.g., Ferguson 1983) and for the

Coast Salish in particular (e.g., Schaepe 2006). For some, like Codere (1950), the presence

of warfare in oral histories apparently would be understood as part of a game, as if

readily converted into “fighting with property” in the elaborate potlatches of the

colonial period. However, as Ferguson (1983:133) pointed out, “Northwest Coast

warfare was no game.” Warfare in the region involved “Sneak attacks, pitched battles,

ambushes, prolonged attritional campaigns, treacherous massacres, sporadic raiding––

these were facts of life from before contact to ‘pacification’ in the 1860s” (Ferguson

1983:133).

Ethnographers first conducted their research decades after this “pacification”

that Ferguson refers to, within a colonial context. Moreover, ethnographers collected

information on matters that they were interested in, and sometimes they did not collect

information about warfare. Even Suttles (1990b:152), who did research warfare (i.e.,

Suttles 1951:319-324; ), remarked that “warfare, or at least the threat of warfare, may

have played a greater role in the development of Northwest Coast institutions than I

–– 99 ––

was once inclined to believe.” Moreover, when warfare has been studied, it has been

treated as a separate topic––optional to cover or not––rather than treated as integrally

related to Coast Salish culture.

In this chapter, I briefly cover the weapons used in the practices of warfare. For

the most part, weapons comprise the tools and material capital used for warfare, or what

could be called the means of destruction––or even further, the means of defense used by

people to preserve their means of production and livelihood. In the next section, I

describe the relations of warfare, the sociopolitical roles involved in warfare, both in

offense and defense, and the nature of organization and authority, or the relations of

production. After discussing the various roles, I discuss the protocols available and the

array of cultural practices that order or structure the nature of conflictual interactions

during the postcontact or ethnographic period. In the last section, I discuss the

motivations for warfare from materials and territory to prestige. In so doing, I continue

the materialist frame and discuss the field of warfare showing how violent practices are

employed to gain various forms of capital.

Before I begin this summary of ethnographic literature in the Coast Salish region,

it is important to keep in mind that the resources available are drawn primarily from

ethnography, historic accounts, and oral histories, each of which has its advantages and

disadvantages. Each of those disciplines has its form and perspective to consider. From

ethnography, we benefit from sustained analysis that brings out insights amidst the

great complexity of cultural experience, but which are distorted by the postcontact

context, as is especially true with warfare (Ferguson and Whitehead 1992). From

ethnohistoric accounts, we gain greatly from the direct (or indirect) observations of

actual events. These often represent a third-party perspective other than that of the

culture under study, however, those events are often described in a way that reveals an

ignorance of cultural practices, while also suffering from the same colonial effect of the

observer upon the observed. From oral histories, we are aided, by the emic perspective

of those informants who have inherited or learned these traditions, and offer a telling

–– 100 ––

that highlights what is important about that event, the protocols enacted in those events,

and even the telling of the tradition itself. Of the three, the sources for oral histories are

of greatest antiquity. However, the greatest advantage of oral histories is also a major

limitation: by gaining insight into the cultural perspective of what has been highlighted

and handed down through generations, we cannot read what is not said, those subjects

or details that might relate to our inquiry but simply were not included as part of their

telling, perhaps considered unimportant or common knowledge to Coast Salish

themselves. Having noted these qualifications, however, the values of each outweigh

the disadvantages, and combined these provide a more well-rounded record. These

accounts will be integral for undertaking an archaeological exploration of warfare.

Weapons, or the Means of Destruction

One class of Coast Salish weaponry included clubs, axes, knives, and spears.

These are referred to as “melee” weaponry, because they are used in the melee of hand-

to-hand combat. Another class of weaponry includes “projectile” weapons which can be

thrown from a distance. These also include spears, but also slings, darts (with atlatls),

and the bow and arrow. Increasingly, during the colonial period, iron axes and muskets

were used. At the surface, these means of warfare are mostly the same as the means of

productive subsistence. For most cases, that is true––these implements are

multifunctional; this has always been the case throughout world history, with peasant

armies formed armed with scythes, axes, and machetes. However, it is also true that

some weapons were distinct from their subsistence form, having different shapes or

decorations if used in warfare, as will be discussed below. Smith (1940:163) conveyed

the significance of their differences among the Puyallup-Nisqually:

The weapons employed in all forms of manslaughter were strictly identified with the killing of humans and were used for no other purpose. So true was this that if a war club or dagger accidentally fell from a man’s clothing during a social gathering, it was understood that he had intended slaying his host or guest, as the case might be, and feeling against him was as strong as though the deed had been attempted. Warriors were fully equipped with the paraphernalia of war but ordinary men frequently owned neither war club nor dagger so

–– 101 ––

closely associated with the shedding of human blood.

Much detail on these tools, their material for construction, and so on, is widely

available in the regional literature, so in the following there is no need to be exhaustive,

however, it is important to lay some groundwork for this discussion.

Melee Weapons

Knives would seem to be an item of dual use, however, one of Suttles’ (1949

[6]:87) East Saanich informants stated there were two different kinds of knives. Some

knives, such as slate fish-cutting knives , would not be suitable for war, although others

might. The Twana had double-edged, chipped-stone daggers of “flint” or obsidian with

bone or wooden handles; no other materials would be used for such blades (Elmendorf

1960:471; Barnett 1955:268-269). One informant described that a warrior’s knife was of a

post-contact trade item, a 10-in, double-edged iron knife with a brass handle (Suttles

1949[6]:87).

Perhaps the most distinctive weapon was the war club. Although some types of

clubs were used to dispatch fish like halibut or sturgeon, war clubs had a specific use

and term. Duff (1952) noted that war clubs were preferred weapons, and they were

often shaped just like sturgeon clubs, although still distinctive. They could be of stone,

hardwood, or bone. Some hardwood clubs were the length of baseball bats, albeit

carved with sharp edges, while stone clubs could reach two feet in length. One of Duff’s

(1952:60) informants related that “these clubs were only brought in at ‘big times,’ and

their histories were told.” Among the Klallam, war clubs were predominantly of elk or

whale bone, over a foot long and equipped with a wrist strap (Gunther 1927:268). No

matter the material, it seems, they might exhibit anthropomorphic faces or zoomorphic

imagery. Among the Twana, the “commonest” weapon was the club, often shaped like

a paddle. Elmendorf (1960:471) noted that for winter dancing, a smaller paddle-like club

might be used as a symbol of a person’s war power.

Somewhat intriguingly, Barnett (1955:269) remarked that the spool-shaped hand

–– 102 ––

maul was used for close-range combat as a minor extension of the fist. He described the

weapon as “embellished with the nipple on the cap at the small end.” This corresponds

to the classic description of the hand maul first noted in the Marpole era and is usually

associated with woodworking activities, which seems more likely the case.

Projectile Weapons

As Elmendorf (1960:470) noted, the spear was a warrior’s weapon—and this

makes sense since its use in hunting had probably long since declined with the advent of

the atlatl propelled dart and the bow and arrow. Duff’s (1952:60) informants mentioned

the use of spears in bear hunting, and how spears were useful as hiking staffs. In

warfare, spears were primarily used for thrusting or stabbing, rather than thrown, and

were tipped with points of bone, stone, or mountain goat horn. A favoured tactic with

spears, when an attacker advanced, was to plant the handle into the ground and impale

the on-rusher upon it (Barnett 1955:270). Among the Klallam, shafts of spears were

made of yew, and could be “two fathoms long” with a large spearhead at the tip

(Gunther 1927:268).

Slings were also used in warfare, made of various animal skins and cords. Using

“perfectly round” stones, the slingshot was swung and whipped around the head.

According to one of Duff’s (1952:60) informants, “One man from Yale was said to have

been able to split enemy canoes with stones up to 4 inches in diameter.”

The bow and arrow was the favoured weapon for long-range combat. Typically,

the bow was made of yew wood (“white cedar”) or vine maple and the bowstring was

made of deer-sinew, or occasionally, sea lion gut obtained in trade from the Penelakut

(Suttles 1948 [1]:84, 1949 [6]:62-63). The bow was held horizontally, rather than upright

or vertical. The center of the bow sometimes was constricted with the tips recurved and

those were “often decorated” (Duff 1952:59). Some bows were also backed with sinew,

attached with a fish-skin glue (Elmendorf 1960:87). Barnett (1955:100) remarked that

there was no difference between hunting bows and bows for war.

–– 103 ––

Arrows had shafts of cedar, with the tail bearing two duck feathers opposite each

other (Duff 1952:59). Their length was the distance from shoulder to fingertip and those

would be sanded smooth with dogfish skin (Barnett 1955:101). In some areas, as among

the Nisqually, arrows had fore-shafts at the end and their arrowheads were often tied on

with cherry bark (Smith 1940). Arrowpoints were made from bone, wood, shell, ground

stone, or chipped stone, and later iron. One of Duff’s (1952:59) informants stated that

the chipped stone point was exclusively for war. Similarly, in Puget Sound, it has been

recorded that there were two types of arrows: one type for hunting and another for

warfare (Haeberlin and Gunther 1930:26). Eells (1985:149) noted that in the Klallam area

arrowpoints were fastened loosely, so that after impact it would be “remaining in the

wound” (see also Gunther 1927:268).

Some stone materials were sought for their magical or “poisonous” qualities. For

instance, black obsidian among the Twana, was held to be “naturally poisonous” inside

the body (Elmendorf 1960:90). Among the Nisqually, quartz points, or “yellow” points,

were also thought to be toxic once they penetrated a person, as were points made from

human bone (Smith 1940:296). Also, natural poisons were applied to arrowheads as

well, as the Snoqualmie used rattlesnake venom caught in the mountains (Tollefson

1996:155). Among the Twana, arrows and spear points were fire-heated to increase their

potency (Elmendorf 1960:471). Sometimes poisons were applied for the same reason.

Among Stó:lō groups, Duff (1952:59) recorded that “war-points were poisoned by

dipping them in human brain,” while Elmendorf (1960:471) found that knowledge of

arrow poisons was a closely guarded secret.

Although harpoons are typically associated with sea-mammal hunting, Gunther

(1927:268) noted that the Klallam would use them in warfare when necessary. Two-

pronged harpoons would be thrown at an enemy: “The points are barbed and attached

to a sturdy rope. When the points have pierced the enemy he is dragged toward the

attacker by means of rope, then clubbed, and his head cut off.”

This is yet another case for the multifunctional and creative use of tools. The

–– 104 ––

weapon categories described above are simply the types commonly described, but even

their construction and materials exhibited a range of variability, especially as described

for clubs, knives, and so on––a variability that seems particular to distinct Coast Salish

groups or simply to individuals. There are also other weapons described by some that

seem rare or as individualistic as Barnett’s description of the hand maul as a weapon.

For instance, Duff (1952:5) recorded a description of a “sort of cross-bow,” which had a

trigger and was used backed against the shoulder; Barnett (1955:270) mentioned that a

chief of the Sliammon had described “a sort of catapult for discharging spears which

was used in attacking a stockade,” made from “a springy sapling of yew.”

Equipment and Appurtenances for Warfare

Armour

In addition to weapons of offense, warriors would also have items for defense, or

usually so. Many mentioned “buckskin shirts” from elk or deer (Suttles 1949 [6]:79),

although it was just as common for a Coast Salish warrior to report that no physical

armour was needed: “Each informant, when asked about armor, immediately replied

that a man's power was his protection” (Smith 1940:164). When Frank Allen was asked

about buckskin or other armour, he replied:

That’s no good. People who use that (armor) are no real war men. Warrior doesn’t care if he dies or lives! Disgrace for a man to fight with any protection but his power. Skokomish, Klallum have got big heart, don’t need that kind of stuff! (Elmendorf 1960:472).

However, even the buckskin hides as armour would allow some flexibility in

movement, as they were sleeveless and hung down to mid-thigh. Rod or slot armour of

the northern groups was not used, as “Men preferred to rely on their agility in dodging

or running” (Barnett 1955:270).

–– 105 ––

War Canoes

Besides the weapons used or armour worn (or not), the means of warfare also

included larger, structural items, such as canoes, fortifications, and other constructions.

The Coast Salish made several types of canoes, predominantly the dug-out cedar canoe.

For rivers, they carved shovel-nosed river canoes that were variably long but typically

narrow and were poled, rather than paddled (Elmendorf 1960:170-76; Collins

1974:64-65). Another typical type of canoe, the “all-purpose canoe,” was carved from

half a cedar log and about 4 to 9 m (15 to 30 ft) long with the ability to handle up to ten

people. The hull is flatly curved, cresting slightly to a point at the stern. Duff (1952:52)

considered this the base Coast Salish type of canoe with variants in length. But, for war,

larger canoes were used, such as the “Nootka type,” a longer canoe with the “wolf’s

head” bow that could carry more people and was used, for instance, by the Stó:lō for

trips to saltwater (Duff 1952:51). These were painted red on the interior and black on the

outside (Collins 1974:65; Haeberlin and Gunther 1930:34). These war canoes could hold

from six to fifteen people (Waterman and Coffin 1920). There were lengthy notches from

the bow that allowed placement for spears or harpoons, and these notches were filled

with shredded cedar bark to prevent loose rattling (Elmendorf 1960:171-72). It is clear

from Duff’s (1952:51-53) informants that the canoe––while named the “Nootka type” (or

“Chinook” by Waterman and Coffin [1920:13])––was carved by Salish artisans, although

it was often purchased; Gunther (1927:212) noted that such a canoe could be traded from

the Nuu-chah-nulth for a slave and would be highly valued.

The Coast Salish were familiar with the Northern style of canoe, called “double-

enders,” with protrusions on both bow and stern (Duff 1952:53). These were sometimes

purchased or otherwise acquired from northerners. Barnston (1998:34), in August of

1827, described the canoes of some Coast Salish groups as they passed Fort Langley to

go to their salmon fishing camps. I provide it at length as he detailed the numerous

styles of canoes used and it also conveys the ways used to carry an immense amount of

goods, a luxury not afforded to most hunter-gatherers travelling across land:

–– 106 ––

Saturday 25th [August 1827]. All Hands employed as yesterday, sick list the same.... Families from the Sanch Village at Point Roberts have been passing in continued succession during the day all bound for the Salmon fishery. Their Luggage as well as that of the other tribes is transported up and down the River on Rafts, which are formed by laying Boards across two or more Canoes Kept 8, 10, or 12 feet asunder. They have also among them large War canoes procured from Indians to the northward, which are used by them as Luggage Boats, and which contain a great Bulk of Furniture & Baggage. The Size of some of these craft is fully 50 feet in length and 6 to 7 ft. across the middle. On the Top of the Stern which is flattish there is in general carved out the resemblance of the face of a human Being, and the Stern [bow] rises to the height of at least 7 feet from the water. Whether this latter be intended merely for ornament or not is impossible to say, but it gives the Canoe an imposing appearance, and must afford to the crew a tolerable defence against arrows when they are advancing straight against an enemy. The Sides of the Bow and Stern are very fancifully ornamented with circles and other regular figures which are laid on with various coloured Paints or Clay (Barnston 1998:34).

Frank Allen, one of Elmendorf’s (1960) Twana informants, had ridden in a

northern-type canoe while on a trip to Fort Rupert, and thought that these did not

handle nearly as well as the “Nootka-type” he was familiar with. On a similar note,

Barnett (1955:114) described the Kwakwaka'wakw canoe, manka, as having a square-cut

hull and a large prow “sometimes eight feet high on which were hornlike projections

and a hooked beak representing an eagle; in a high wind it was removed since it offered

considerable resistance.” Such an artifice, however, was likely left on for battles, to instill

fear and block arrows. This suggests some differences in approach between the two

groups.

As with any technology, there are advantages and disadvantages, and the

ramifications of these would have played to their tactics in battles. For instance, John

Fornsby recounted how his grandfather encountered the Lekwiltok on the open water.

My grandfather, my youngest grandfather, sqáyxe, had gone to see his relatives at Lummi. The yúk'wta [Lekwiltok] tried to catch my grandfather, but he had a fast canoe. They chased my grandfather and his wife, chased them for a long way and tried to catch them. But they never caught them. They gave up. Their canoes were no good, I guess (Collins 1974:116).

According to Arvid Charlie, in his recounting of the Battle at Maple Bay, the

–– 107 ––

Coast Salish canoes were generally smaller and more maneuverable, as opposed to the

larger northern canoe (Angelbeck and McLay 2009). From the descriptions, it appears

that the northern canoe style was better for long distance travel and would look

imposing as it came to shore for a raid, but for sea battles, as at Maple Bay, this appears

to have been a weakness the Coast Salish groups seized upon. Like spirit power (and

nakedness) for armour, this again suggests a Coast Salish preference for flexibility,

agility, and speed, rather than relying simply on mass or force.

Defensive Architecture and Head Poles

The Coast Salish also created an array of defensive sites. The largest involved

palisaded fortifications often on bluff-tops or steep-walled spits. In front of the palisade,

there were often trenches and embankments. In the Fraser Canyon, some sites were

defended with rock-wall fortifications. Other types included refuges, generally in

naturally defensive locales, or lookouts at exposed, elevated positions with views of

travel corridors. More discussion about these sites occurs in Chapters VII to IX. For

now, it is important to note that Coast Salish accoutrements for warfare extended well

beyond weaponry and included multi-person devices like canoes as well as construction

intended for use by multiple people, and requiring the organization and cooperation of

many to build. Besides defensive architecture, there were also other embellishments at

many sites that were related to and a by-product of warfare:

We saw no village nor inhabitants near the place. But on the point of the beach there stood a remarkable High pole, strongly supported by props at the Bottom, and at the top of It was fixed a human skull. What the reason of so curious a thing could be no one could divine. Many such had been seen in different parts of the Inland Navigation and in Mr. Hanson's late cruise. No less than three of these Poles with skulls on them were seen at one place contiguous to which was a very large burial ground.... ––Anonymous member of Vancouver’s expedition, 1792 (cited in Collins 1974:28).

Many encounters detail the impaling of enemy heads upon poles outside of the

village. Boas (1889:324) described that among the Snuneymuxw, “The heads of the slain

–– 108 ––

were cut off, taken home, and planted on poles in front of the houses.” Barnett

(1955:269) similarly described these heads “unceremoniously stuck on poles which were

planted on the beach, usually at some distance from the village.” At the head of Howe

Sound, in Squamish territory, it was noted that a place called Whoh-nuck, separate from

the village, was where the heads were displayed (Matthews 1955:429-430). Suttles

(1990a) remarked that these represented “trophies.” But, these can also be seen as

demonstrations of interpersonal power (ii), subsequently used to convey personal power

(i).

In the story, the “Myth of the Ghost Lover,” retold by Hill-Tout (1907:338), a raid

by the Songhees on the Sechelt was successful in bringing home many severed heads,

which they put on poles. A woman walks by and one of the heads is so handsome, she

takes it down and caresses it for many days. Eventually, the ghost of the head starts to

speak to her and begins to repeatedly visit her the following nights. For the purposes

here, the story suggests just how widespread and common the practice was.

Warriors

As he was dying he spoke to his people, telling them that he had not become a warrior simply to make himself “big among his own people” but had done so because he had been ashamed that the northern people had taken their children as slaves; he had become a warrior in order to protect his people.––Wayne Suttles (1951:324)

We change our voice now to talk about warriors.––Frank Allen (Elmendorf 1993:126)

The Northwest Coast has long been known as unique for its hunter-gatherer

societies that exhibit an array of specializations more typically associated with larger-

scale societies such as chiefdoms and states. There were shamans, but also carvers,

carpenters, orators, and others––some hired for their particular skills on occasion. There

were also professional warriors. As Barnett (1955:267) described, warrior status was a

“professional one and ran in families. A northern father tried to inculcate the desire to

–– 109 ––

fight in at least one of his sons.” Each community would have a warrior or two; Suttles

(1951:323) stated that the Lummi may have had four, and Frank Allen also mentioned

that the Snohomish had four warriors at one point (Elmendorf 1960:467). They could be

hired for particular reasons as well, as one Skokomish warrior was hired to come and

kill a troublesome bear, “which he did singlehanded, using only a knife and his war

power” (Elmendorf 1960:467).

It is not that these were the only fighters to go on attacks or defend the village,

rather these warriors were the ones who led such endeavors. And they trained heavily

for the ability to do so, in order to acquire spirit powers for the role, just as other

professions had their associated powers. Warrior powers, however, required extensive

training, much like a shaman did, which would keep them apart from the community

for extended periods of time. For this training they might have to submerge deep to the

bottom of lakes (Suttles 1948[3]:63), or engage in other difficult travails.

Among the Twana, these powers they acquired were described as “bad” spirit

powers, or warrior powers, called sča'laq (Elmendorf 1960:467). Some might attain a

wasp power, yellow-jacket power, wolf power, and among the Sliammon, the double-

headed snake, sisiutl, was considered quite powerful (Kennedy and Bouchard 1983:90).

Another power widely sought was thunder power. A Lummi informant told Suttles

(1949[5]:90) that one who got the thunder power “couldn’t live with people. Built house

way upon top of hill.... Had to keep eyes shut up[,] opened like lightning[,] like

thunder.” Among the Klallam, an informant stated that “Thunder [was] used mostly for

war power. When warriors get into canoes ... talk of war, thunder rolls” (Suttles

1952(13):81). Those warrior powers were so great that once acquired, they were

supposed to use these powers to help the others. As Julius Charles emphatically told

Suttles (1948[3]:64), “What he gets has to protect his people.... You write it down there.”

No doubt partly because of these “bad” powers, there was an ambivalence

regarding warriors. As Barnett (1955:267) described, “in every village there were men

who were called ‘mean’ .... [and] they were animated by horrendous spirits.” They

–– 110 ––

sometimes lived away from the village in isolated houses. For the Sechelt, Peterson

(1990:38) described this dangerous aspect of warrior, or sky'akth, as one “dwelt apart

from the village, with no fixed abode; and children were forbidden to go near him, lest

he harm them.”35 Florence James, a Peneklakut elder, described the Cowichan warrior,

Tzouhalem, whom she acknowledged: “Every one knows him as a bad man ... and as a

man like a monster. But, to me he was a warrior” (Angelbeck and McLay 2009).

In a way, these warriors are quite similar to shamans; in fact, a form of structural

opposite, with one operating by predominantly physical means, the other

metaphysically. While shamans, also called “doctors,” are able to heal the wounds or

sicknesses of others, restoring a person’s health, warriors inflict wounds and kill

others––they were agents of entropy. Shamans could also be or act in a negative mode,

conjuring sorcery upon others for ill (a reason to be wary of their dangerous powers as

well). One of Suttles’ (1950 [9]:111) informants, Julius Charles, stated that “Long ago

only doctors killed each other,” presumably indicating a time before widespread

warfare.36 But warriors also exhibited a positive aspect, as the agents of preservation

and defense for the community.

Another duality involved the direction of energy flow: as shamans typically

fasted for their powers, warrior powers (at least once acquired) were associated with

supernatural abilities of consumption. Smith (1940:74) described how Nisqually

warriors consumed more, but did not show it physically, as if their spirit powers

consumed it. There were accounts of warriors eating a side of beef or drinking a full

barrel of water––this expresses a lack of restraint, as opposed to what chiefs convey.

Twana warriors, while they did not behead their enemies, also drank the blood of

35. It has been noted that contemporary veterans among the Stó:lō have been similarly treated with a form of wary ambivalence, leading to a lack of support when returning from war, an effect of the lasting sentiment that they bear strong or dangerous warrior powers (Carlson 1997).

36. “The shaman got money for killing somebody.... These shamans sang their power songs in order to kill somebody. They would help a sick person get better, and they would get money from the sick person. Anybody might be killed by the shaman, and then the shaman was killed” (Snyder 1968:97).

–– 111 ––

enemies or ate a bit of flesh of the dead enemy, in order to “feed” his power. Elmendorf

(1960:469) went on to say: “war powers liked blood or the flesh of enemies. When not

on raids warriors drank animal blood from time to time for the purpose of placating this

sanguinary appetite of their guardian spirits.” During spirit dances, warriors might cut

their own flesh and drink the blood (Elmendorf 1960:467).

Perhaps it is this opposed quality that leads to some shared traits, as warrior

spirit powers enabled them to heal their own wounds. Moreover, whereas most

individuals’ spirit powers––that of the winter dancers, for instance––were only available

in winter, when the spirits came down from the mountains––the shaman’s and the

warrior’s spirit powers were the only specializations with access to their powers year

round (Collins 1974:118).

To counter Western conceptions of a soldier as purely a physical agent or ground

troop, it might be better to conceptualize the Coast Salish warrior as a “warrior

shaman,” highlighting the importance of spirit powers in his ability to attack and

defend. In one account, by John Fornsby, a Skagit warrior named Old Snatlem used

only his strong warrior spirit power to defend his village from attack by the Lekwiltok.

He drove the winds and made the waves crash high to prevent the Lekwiltok from

coming to shore. Said Fornsby, “Nobody was killed that time.... Old Snatlem was a

powerful man. He made it blow hard” (Collins 1949:299). Chief Sampson of the

Swinomish recounted one story about a retaliatory attack on a Skagit group at Ut-sa-

laddy, where a warrior used his power and “The Skagit fell dying in their tracks and [the

warrior’s] older brothers and father finished them off with war clubs” (Sampson

1972:57-58).

Shamans also could conduct warfare, albeit restricted to battles of sorcery. It was

not uncommon for odd deaths to be attributed to the malignant powers of a shaman

even from afar. This was referred to as “power shooting” and “soul theft.” Once a

person’s spirit was taken from a doctor, they were not fully present anymore; as Frank

Allen described the victim in one retelling: “He’s in the ghost land. That λ'pα'xcut has

–– 112 ––

taken him to the ghost land” [Elmendorf 1993:224]. A battle could take place in the

spirit world as two shamans battle over the spirit power of a person. Stories of the

“power wars” are told to this day in the Upper Skagit area:

Larry Williams: They used to tell me stories of them old Indian doctors, the wars weren’t just physical, but they were power wars too. Bill Angelbeck: What do you mean by that?Larry Williams: Spiritual power. And, that’s why I was saying that, you know, throwing power, and that’s why nobody would come up here because those Indian doctors they would throw things at them if you would.Sherman Williams: They kill one another that way. No one believes us, but we’ve seen those days, you know.Larry Williams: So, [the] kind of powers [that] used to reside up here––and that’s why this [area] was protected that way (Miller and Angelbeck 2008b:113).

In another oral history, a shaman by the name of Little Sam killed another

shaman by “hanging a rush effigy of this shaman’s spirit on a house post. The following

day the people found that the shaman had hung himself” (Haeberlin and Gunther

1930:78). Often, a shaman after healing a victim would be hesitant to announce who had

caused the spirit-loss or sickness of a patient; as Henry Allen stated: “he’d get paid for

the job, and it might be himself next time!” (Elmendorf (1960:510). A shaman might be

hired as well to kill “a warrior that had become dangerous and overbearing” (Haeberlin

and Gunther 1930:78). Curiously, while most shamans were healers, “a shaman could

not help a person wounded in war” (Haeberlin and Gunther 1930:78). Warriors had

spirit powers (or were left) to heal their own wounds.

A shaman not only would attack one person or another shaman, but could also

attack whole communities. In one account, a Skokomish shaman sends his otter power

in attack upon a Skagit village and their houses are devastated in a storm; “his doctor

power killed them” (Elmendorf 1993:163). Elmendorf (1960:510) remarked that these

feuds of hostile magic became “a recognized substitute for open hostility in the form of

retaliatory raiding.”

–– 113 ––

Leaders for Defense

While warriors might be considered opposites of shamans for their powers to

inflict death or provide defense, or protect life, a warrior might also be considered the

obverse of a chief in the political realm. As Edmond Lorenzetto told Wilson Duff

(1952:82), a warrior is “almost like the opposite of sie'm.” Among the Coast Salish, the

sie'm or chief was a person regarded for peace. One of Collins’ (1974:36-37) informants

stated that “the chief told people to stop when they wanted to fight. The chief didn't

lead people in battle, didn't know how to fight” (emphasis added). Chiefs, in part, often

attained their status as a leader through mediating disputes. They respected or were

given the authority to help resolve the dispute for those involved. As Miller

(2001:149-150) detailed, the powers of chiefs enabled them to settle disputes, and they

were often called on, even hired, from abroad to provide their “good talk.” Their power

was shown in being able to mediate between distinct domains, such as two parties in

conflict. Miller (2001:115-116) emphasized that the power of a chief came not from a title

or role that they played, but rather resided in the person. Because of their ability or

powers (usually synonymous) and the respect they had earned, they deserve or have the

right to be the chief. The chief’s status and power was based in the individual, not in an

institution or role.

Collins (1974:114), in the following passage, reveals the contrasting relationship

between chiefs and warriors:

...war leader [was] also distrusted. Warriors per se did not have the prestige which they had among the Plains Indians. The same qualities which made a man a good warrior––a hot temper, an indifference to personal risk, willingness to inflict physical injury––were at variance with the image of the ideal man as slow to wrath and hesitant to strike another.

There are also similarities between the actions of chiefs and shamans. While the

shaman, as “doctor,” physically restores health to an individual, the chief politically

restores stability to disruptions in the community. Warriors can be seen as the negative

application of this role, as they attempt to maintain the health of the community through

–– 114 ––

active and often destructive defense. Professional warriors would actually be given

authority during times of war, albeit only for the duration of the battle (Suttles 1951:277).

According to Smith (1940:156), chiefs would consult with the warriors about the

situation, but “the warriors were not placed in charge until the moment of actual

fighting” lest they start the fight because they want to fight––as if unleashed. It is in this

sense, that warriors were seen to be, as Smith (1940:51) phrased it, “necessary evils.”

During peaceful times deference to the authority of the warrior was tinged with fear and, because of the general repugnance for open conflict, every effort was made to keep him from displaying force. But during periods when peaceful means no longer served, he was given, in his role as an expert in war, rather complete control of the situation (Smith 1940:50).

The chief, in a sense, needs the warrior. In this complex hunter-gatherer society

with surpluses aplenty generated through their productive subsistence and surplus-

generating activities, a means to protect those surpluses or their means of production

(weirs, reef-nets, clam gardens, etc.) was “necessary.” As Smith (1940:50) noted, “The

warrior was important because he protected the interests of those around him from

threats of violence, threats which might at any time disrupt economic activity.” Gifts

also were often presented to warriors to prevent any violence.

However, the Coast Salish are variable across the region, exhibiting many local

differences, and so, for instance, professional warriors sometimes were noted as

household leaders as two Skokomish chiefs had been, according to Elmendorf

(1960:473), although he remarked it was atypical. Even for “village leaders” Elmendorf

(1960:473) commented that this was indeed rare: “A Duckabush warrior was at one time

leader of that Twana village, but this is the only instance in my data of any village

headman being a warrior.” Chief Seattle, according to oral tradition, rose in power by

leading a defense against an attack from the east (Costello 1895). Even so, that is also

qualified, as his leadership of the Duwamish and Suquamish was also due to his great

skill in oration, as he is often quoted in relation to peace. Another example of an

exception to this warrior discussion would be the Northern Coast Salish, such as the

–– 115 ––

Comox or Sechelt, where Barnett (1955:267-68) described that spirit powers were less

needed; among the Sechelt, it was simply a secular form of training, meaning that it did

not require warrior spirit powers. While the Comox had warriors as a part of sodalities

or secret societies, such as hamatsa, as did their northern neighbors, the Lekwiltok, with

whom members would often ally. Among the Quinault, there were no roles for the

warrior as a specialization; all able males would just fight (Olson 1936).

Roles for Defense

Besides the role of the “warrior” as the organizer of resistance against raiders,

there were other roles involved in coordinating defense across the Coast Salish region.

These commonly included scouts: “When the Snohomish heard rumors of an enemy’s

approach, they sent young men as scouts” (Haeberlin and Gunther 1930:13). When the

attack did happen, for instance against the Semiahmoo, a runner from the Semiahmoo

would be dispatched to the Lummi (Suttles 1951:322). In other areas, messages of

attacks would be conveyed by watchmen at “fire signals.” The Nisqually would have

people manned at “fire signal stations at various points. The last of these posts, (tatu’so),

was situated near the present Tacoma Hotel” (Haeberlin and Gunther 1930:13). Closer

to the villages or defensive refuges, lookout stations would also be manned. Among the

Sechelt, a “lookout tree” would be manned near their palisaded fort (Peterson 1990:28).

The Quinault, in times of perceived threat, would place sentries along a high platform

that rimmed the palisade of their villages; they were known as suxwanaxwame'ana, or

“they who watch” (Olson 1936:117).

During periods when tensions were high, as around 1850––when the Klallam

were warring with the Cowichan––the village ensured that watchmen near their Port

Townsend area village would be vigilant: “Every day and night they watch over on the

spit outside Port Townsend, watch for those Cowichan to come,” Frank Allen recounted

(Elmendorf 1993:133).

Among the Lummi, the warriors were assisted by young men, or those less

–– 116 ––

experienced with battle, in the canoes. The warriors would position themselves in front,

ready for attack, while younger fighters would paddle behind (Stern 1934:98; Suttles

1949[5]:69). Also, elderly men often would join as well but to sing their powers to aid

the warriors (Stern 1934:98).

Women and Warfare

While most of those involved in warfare as warriors or fighters were male, there

were roles for women in conflict as well. As Suttles (1951:323) noted, “Usually everyone

in a community helped defend it….” Able women were likely leading retreats into

hidden refuges behind villages or underground houses, protecting the children and

aiding the elderly. In one important way, women often gained clairvoyant powers (Hill-

Tout 1978 [1907]:162), and these could serve well in defense. The spiritual power of a

seeress could alert the villagers of coming raiders, much like a shamanistic scout (Suttles

1951:322). When the warriors left in times of war, whether in raiding or retaliation, the

wives of the warriors would sing songs to support their husbands from afar. Among the

Nisqually, Smith (1940:165) described the implications of this singing:

Wives or female relatives of men who were absent, whether on war trips or on hunting, fishing, etc., expeditions, sang a song which brought their men good luck. This song was said to have been very beautiful and, although it could be sung upon social occasions as well, its main purpose was as described. During it the women faced in the direction which the men had taken. When it was finished, if one woman started to cry it was known that her husband or relative had been killed.

Before an attack on a home village was to begin, women beat sticks against their

houses to “stimulate possession in those warriors who had spirit power” (Barnett

(1955:270). In other ways, women would also use their songs to help stimulate the

warriors during battle, standing behind the men. According to Suttles’ (1951:322), one

woman was known to sing behind the warriors and use her power to “dull the senses of

the enemy.”

A woman’s power might be more than sorcery in battle. Some women, among

the Upper Skagit would travel in raiding canoes. Collins (1974:115) remarked that the

–– 117 ––

“[raiding] party left the few women who accompanied them to watch the canoes” while

they attacked. There are also accounts of women as warriors. A woman warrior that led

a routing of the Cowichan during one attack. As Frank Allen described, among the

Klallam, “A woman is there named sαkema’yIł. . . . Nothing can catch her; she runs like a

wolf” (Elmendorf 1993:133).

Sαkema’yIł, along with a male warrior named xe’tanǝxw, headed up a bluff in the

expectation of a retaliatory attack by the Cowichan. When they spotted the Cowichan

approaching, she told her partner “Wait a minute, till they’ve got down under the bluff.”

The Cowichan then begin to approach the village below them. She then tells xe’tanǝxw:

“I’m going to shoot! You holler your tamánawis now!” referring to his spirit power.

Frank Allen then said:

So she shoots and kills one, and xe’tanǝxw hollers and runs after them. They think lots of Klallam are coming after them and they break and run to the spit where they landed and had a watchman at the spit.sαkema’yIł stops and cuts the head off the man she has killed. That’s her game. She was a great woman. I saw her in my time, a small lady, not big (Elmendorf 1993:133).

Organization of Defense

Most chiefs’ authority, in Coast Salish villages, generally extended to the limits of

their household. As Suttles (1951:277) described, while one might be seen as a village

chief, he more operated as a “potlatch organizer” for the village as a whole.

Economically, households functioned autonomously for most tasks. But, when it came

to the defense of the village, Suttles (1951:277) noted that the villagers generally worked

together: “The village usually, though not always, functioned as a unit in defending

itself against enemy attack. And the village might function as a unit in potlatching. But

there were probably no other functions of a village as a whole.”

When it came to defensive fortifications, the whole village might work on the

fort’s construction, as Julius Charles related to Suttles about the fort at Gooseberry Point

(1948[2]:83). So, just as the warrior’s authority over the village is temporary, during

–– 118 ––

times of attack, so is the operation of the community as a whole: it is done for defense, or

perhaps for potlatches. However, Barnett (1944) described how underground dwellings

for defense would be built by families, especially among the Northern Coast Salish,

although while autonomous in construction, there might have been coordination still at

larger scales, as Schaepe (2000, 2001, 2006) has argued archaeologically for the rock-wall

fortifications in the Fraser Canyon. This kind of coordination is a topic that will be dealt

with in more detail in Chapter VII.

Warriors and Fighters

In the literature of the Coast Salish, warriors have been discussed in a somewhat

conflated manner. Above, I have descibed how professional warriors were given

temporary authority to defend the village, however, these warriors also sought the help

of all other capable and willing males to be defenders, or perhaps assembled a team for

an attack. As fighters, these are not “warriors,” in the Coast Salish professional or

specialized sense. Here, it is useful to keep the categories separate. Maintaining this

distinction between warriors and fighters perhaps offers a way to interpret the conflated

manner of how “warriors” are occasionally described. For instance, warriors are

described as undergoing rites of purification; Barnett (1955:269) mentioned that “the

returning warriors of the Klahuse, Sechelt, and Squamish went through a ritual

purification,” involving practices such as bathing, bleeding, and enduring sweat baths.

Similarly, for the Twana, Elmendorf (1960:470) described that “Before actually entering

his home village each member of a returning war party underwent ritual purification

paralleling that of a homicide or a corpse handler.” The purification might keep those

warriors away from the village for up to four days, bathing and scrubbing with cedar

boughs. The intention of this purification ritual was, in the words of Frank Allen

(Elmendorf 1960:470), “to take away the blood (of the slain) from their bodies and

everything bad.” Rather, when viewed through that distinction––in which informants

made between warriors and other male defenders or raiders––fighters who only

–– 119 ––

temporarily engage in the “dangerous” practices of warfare would need purification to

return to the more optimal settings of stable life in the village. Warriors sometimes were

even described as solitary, having to live apart, for example, on a mountain top, as an

above story recounted, lest the warrior’s power––thunder––strike someone with

lightning when he opened his eyes. Warriors drink the blood of their enemies; fighters

scrub the blood from battle with rituals akin to that of handling corpses. If a young

fighter had aims of becoming a chief, these dangerous and ambivalent aspects of warfare

would need to be cleansed from him.

As rituals of purification, this suggests the procedure likely followed the general

course of rites of passage. These individuals are temporarily transformed into warriors.

As van Gennep (1960; see also Turner 1969) had outlined, rites of passage proceed in

three stages: separation, liminality, and reaggregation. Eliade (1959) emphasized that

most rites involved the separation from profane world into the sacred. To enter the

sacred, the rite creates a separation from the profane world to temporarily enter the

sacred, and then a reaggregation into profane or routine life. This moral distinction in

the sacred and profane seems not to describe the Coast Salish conception accurately, as

warfare involved the profane, or sacrilegious, the dangerous and ambivalent “necessary

evils” of preserving one’s lifeway––it is liminality that gives its ambivalent qualities. For

such reasons, fighters required rites of purification to return to normal life.37 For the

Coast Salish, in any case, the act of battle and killing often was purified from them

ritually.

If rites of purification were used for returning fighters, then likely there were

rites for entry into the mode of warrior. Perhaps the war dance served this purpose for

some Coast Salish groups, as Samuel Hancock (1927) had witnessed a war dance on

37. This need for temporary separation has similarities to the hunter tradition among the Navajo,wherein the qualities of the hunter are dangerous and predatory, and requires a ceremony for transformation into a predator, and then a purification to return to humanity. Luckert (1975:149) argued that it was the association of guilt with a hunter for killing its prey; thus the rite absolved them of that guilt––it was not even the hunter but the predator they had transformed into (usually a wolf) that had committed that act.

–– 120 ––

Whidbey Island (see page 79). Smith (1940:165) stated that the Nisqually engaged in

such preparations, which she termed not a “war dance” but a “power sing”: “Before a

battle or going out from the home village the warriors engaged in a power sing to call

their powers to their aid.” Often, before battle, warriors and fighters applied the

ornamentation that symbolized warfare. A warrior’s face would be painted black, with

a concoction of “soot and grease” (Elmendorf 1960:472).38

Leaders for Offense

In the south (Nanaimo, Cowichan, Sanetch, Squamish, Maskwiam [Musqueam]), the origin [of a raid] was very frequently a desire on the part of a novice warrior to try his newly acquired spirit powers. The young man simply constituted himself the leader of a war party and mustered enough men to accompany him. Usually these were relatives and older warriors, but there were always a number of disagreeable and socially worthless individuals who were glad of an approved opportunity to create terror and to kill and loot (Barnett 1955:267-268).

As organizers of attacks, Collins (1974:114) noted that “a warrior could not force

another to go with him.” If he had lost warriors before, perhaps others would be hesitant

to join their efforts again. But, if successful, there likely would be other raids. To chiefs,

associated commonly with peace, they must have considered raids of this sort

ambivalently or as dangerous, regardless even of spirit powers, given that it would open

up opportunities for reprisal attacks.

To return to Wolf’s (1990) modes of power framework, warriors had innate

power (i), acquired from the war-specific spirit power. They also displayed, in their

defeats of others, power over another (ii), demonstrated even afterward by taking of

heads. However, as leaders for defense and offense, warriors exhibit organizational

power (iii) as well, albeit for a different mode than that of household chiefs. Any

38. Chief Frank Malloway, of Yakweakwioose, described the black paint for spirit dancers as made from the charred remains of any prickly or barbed plant: “The black paint––they usually use devil’s club. They dry devil’s club, and use the cinders; they burn it. They don’t let it turn grey, crushing it all the time, and they sift it, and they mix it with deer marrow. For black paint, they also use stinging nettles. Anything with thorns on it, stinging nettles, beehives.... It’s supposed to be your protection, anything with thorns––it protects you, even if it’s burned” (Angelbeck 2003:60).

–– 121 ––

plunder taken and added to their wealth, would be an expression and verification of the

strength of their spirit power itself.

Protocols of Conflict

Internecine wars are perpetual among the tribes [of southern Vancouver Island]. There are always some old-standing differences between them which are liable, on the slightest occasion, to be revived. Grudges are handed down from father to son for generations, and friendly relations are never free from the risk of being interrupted. Lives taken in one tribe can only be compensated by the same number being massacred in another, and without regard to the guilt of the individuals sacrificed. It is difficult to perceive how, upon such a principle, the extermination of the conflicting parties, eventually, can be avoided.––Matthew Macfie (1973 [1865]:471)

In his classic text, Ethnic Groups and Boundaries, Barth (1969) maintained that the

symbols of a group’s identity were strongest at its boundaries with other cultures. His

message was an important one, emphasizing that group identification intensifies when

confronted with another culture; the corollary to this is that there is less of a need to

present symbols of group identify to others that share your culture. In the Northwest

Coast, however, a different dynamic appears to be at play. Rather, villages and even

households were autonomous, so the zones of contact––the areas where identity is

strongest––occurred even within villages. Contests of identity happen with each

potlatch, each gathering for winter dancing, each time sticks were thrown in slah'al

games, with each stone carried in athletic competitions, and each time disputes arose.

The autonomy of Coast Salish households and individuals led to a high degree of

interactions at the boundaries within their own village and region. This accounts in part

for the immense concern with individual status and the status of one’s household, as

well as the widespread social complexity albeit in uncentralized ways that has always

made the Northwest Coast an exception to existing models of sociopolitical complexity.

The concern for personal image was so central that a blood feud could erupt over

a slight during a potlatch––even when unintentional or simply perceived as a slight by

some. That was reason enough for an attack. In fact, the one who made that slight may

–– 122 ––

have intended for such an outcome. The slight itself was meant to take another’s

prestige down a notch, and there was no better time for such a “stab” than when

speaking publicly at a potlatch or other ceremony. As Barnett (1955:253) described,

“Whenever a slight, whether intentional or accidental, had been put upon an aristocratic

person’s dignity, the response was an immediate reassertion of worth” by a face-saving

ceremony, or through warfare as MacFie (1973 [1865]:471) described.

Much of Coast Salish interaction, conflict or dispute resolution revolved around

notions of the restoration of relations. As Miller (2001) related, the intent of Coast Salish

systems of dispute resolution was to restore relations in the community, not to punish

the individual offender. Those relations, however, do not intend to restore some

utopian or egalitarian sense of communal balance; instead, it is the restoration of

relations, but not necessarily the same relations as before. In some cases, it could be a

restoration from feud to friendliness. But, it as well could be a return to relations

between two groups in which there became an acknowledgment of one household’s rise

in status corresponding to the other’s decline.

Macfie (1973 [1865]:471) had written, quoted above, that the nature of warfare is

seemingly chaotic, “internecine” wars as “perpetual.” Snyder (2002) discussed the

“anarchy” of “primitive warfare,” using the term in its sense of turmoil. Rather, just as

anarchy can also be a reference to social order, these acts of conflict also were actions

taken within a common set of practices. In practice theory, those that are offended can

turn readily to the available practices within a social field, their cultural sphere of

interaction. Accordingly, any individual is Levi-Strauss’ bricoleur, or handyman,

strategically improvising one’s actions or tactics to reclaim or enhance their own status.

These options may involve conflict, or not. Generally, negotiations between the two

parties are attempted first.

Clausewitz (1911) said that warfare is the continuation of politics by other means.

Warfare for Coast Salish individuals can be seen as one of the many options available to

them––not just options to fight or take revenge, but also to pursue negotiation. It is

–– 123 ––

mostly through negotiations of disputes that the restoration of relations are achieved,

resulting in the status quo or with increased power and prestige for some, or even for

the one who resolves the dispute. This type of practice ,where the Coast Salish utilize

the options available to them, as Collins (1979) has written, extends deeply into how a

Coast Salish individual identifies himself or herself. Due to the Coast Salish bilateral

reckoning of kinship, one has options to identify oneself primarily through the mother’s

or father’s line, allowing each to become part of the household that will, in their

estimation, provide them the best opportunities for their own advancement in prestige.

Collins (1979) called these options available in multilineal descent a “Coast Salish

strategy.” This flexibility helps to increase their individual power (i), in Wolf’s (1990)

terms––a similar Coast Salish strategy also is employed for the reckoning of how they

relate to others, or interpersonal power (ii).39

When someone has had his or her status challenged or diminished by insult, or

by offensive actions involving adultery, rape, fighting or murder, the offended party can

request payment, money to “pay” for the damages and cover the offense, whether a

slight or murder. The amount of payment is commensurate with the degree of damage

to one’s status, or equal to the status holder, whether the killing was accidental or

intentional. If the offended party accepts the terms of the payment, or negotiates for

acceptable terms, then the matter is settled. If a slight, a “name saving” ceremony could

resolve the matter; if a murder, the funeral may serve as the occasion to publicly address

how the matter has been resolved. However, as Elmendorf (1960:476-477) has noted for

the Twana, one of the most common reasons for conflict was the failure to pay blood

money: “If offered by the killer’s family and refused, this was an announcement that the

victim's relatives desired to seek blood revenge.” These offers and negotiations for

39. McDonald (1998 [1830]:137) commented on this tactical behaviour of the Coast Salish, explaining why the traders at Fort Langley would not become involved even in attempting to resolve a Musqueam-Kwantlen dispute: “If we did interfere tis possible the parties for the present would acquiesce in our decision; but [it] would only be involving us in endless treaties among them without producing any permanent good––we therefore make it a point to keep Clear of all Indian broils: for, like the generality of their race, they have a happy knack of turning every thing Said or done to their own advantage in Cases of this kind.”

–– 124 ––

blood money were made by a “paid envoy,” that was “not a close relative to either of the

negotiating kin groups” (Elmendorf 1960:477). For the Puyallup-Nisqually, Smith

(1940:155) noted that the groups would meet to negotiate, with help from other leaders:

“If the settlement misfired, however ... it was equivalent to a tacit declaration of war.”

In the case that Hancock (1927) witnessed on Whidbey Island, recounted above

where the Snohomish had kidnapped one of his wives, the Snoqualmie chief did not opt

to negotiate but wanted to attack directly. Hancock encouraged negotiation, but that

was a common course to take. In any case, the chief’s organization of his village and

allies produced a show of force that helped to broker a settlement. Finally, a settlement

occurred, and warfare was avoided.

John Fornsby, an Upper Skagit elder, related to Collins (1974:121-122) a story

about how the Swinomish and Lower Skagit men got into a brawl, ending with a Lower

Skagit man being shot to death. Fornsby said,

The Swinomish gave the Lower Skagit blankets and things––guns. Then they got all right. It was a law they made. The fellows who killed a man had to give something to the relatives of the man they killed. They call this oábilik.40

From these examples, a range of responses is apparent, from straight negotiation

to conflict, with an array of bravado or show of force in between, which in its nature

encourages negotiation through displays of power that indicate the potential or

likelihood of physical conflict. And, as Macfie (1973 [1865]) had surmised, a principle

such as this, can lead to an unending escalation of conflict. He did not, however, seem

to notice how these conflicts could be resolved at various points throughout the back-

and-forth turns of escalating conflict.

Earlier, I described the “Myth of the Ghost-Lover,” a story retold by Hill-Tout

(1978 [1907]:139-142), where a Songhees woman falls in love with a severed head that is

impaled upon the pole from a victory over the Sechelt, although I only described the

40. This is an example of Kelly’s (2004) social substitutability, discussed above, where an attack upon one individual is perceived as an attack on the whole group.

–– 125 ––

premise. The story continued with the Sechelt head, or ghost lover, encouraging his

Songhees maiden to travel to a mountain near Sechelt and visit his brother, who looks

much like him. She does, and they find a way to marry, and the story closes, according

to Hill-Tout’s informants, as the reason there has been peace between the Songhees and

Sechelt ever since, settled by a marriage and alliance.

These options, or cultural practices, are more available to affiliated groups.

These groups are more likely to have intermediaries that can act as third-parties. They

will also likely have relations on the other side that they will not want to have harmed.

Also, they simply have some common interest to draw upon that can lead to negotiated

settlements. Conflicts with more distant groups do not have these types of options

available to either side. This is why Keeley (1996:131) pointed out that there is more

conflict at frontier zones between cultural groups, noting that these interaction zones

“necessarily lack the very social and cultural features that prevent disputes from turning

violent.” In such frontiers, conflict is more readily the option to choose, which would be

the case with distant Coast Salish groups, but even more the case with the Makah, Nuu-

chah-nulth, Nlaka’pamux, or the Lekwiltok.

Regarding the latter, Thom (2005) described how, soon after the Battle at Maple

Bay, Coast Salish groups established marriages with the Lekwiltok, including a key one

between a Cowichan woman and Lekwiltok man from Cape Mudge. As Thom

(2005:362) noted, “This important marriage reopened the Cape Mudge area for island

Coast Salish people to fish and camp at for generations after the couples were wed.” As

Simon Charlie, one of Thom's informants, related:

But that’s when we stopped. They wanted to stop the wars that we had with the Yuqwulhte'x. So they got two young people together to stop the war. Those elders [were] one of the last ones that got married to a Yuqwulhte'x person. They stopped the war. That is why we got that fishing ground right there in Cape Mudge (Thom 2005:363).

Simon Charlie also discussed mountain goat hunting areas they were able to

access in Knight Inlet, even deeper north into Kwakwaka’wakw territory. Thom (2005)

described how these relations have continued since those marriages with the exchange

–– 126 ––

of names and hereditary privileges. These interconnections allowed some Cowichan,

Snuneymuxw, and Comox groups, among others, to return to their seasonal camps or

hunting grounds in the north, access to which had been closed during the Lekwiltok

southern expansion into the northern Gulf of Georgia. The point is that the Coast Salish

resolved this seemingly interminable era of warring by bringing the Lekwiltok into

affiliation.

Still, while there was an array of protocols for resolving conflict, there still

remained the matter of the initial transgression: the unpredictability of slights and

offenses. A variety of motives drove the forces of warfare.

The Forces of Destruction

In the development of productive forces there comes a stage when productive forces and means of intercourse are brought into being, which, under the existing relationships, only cause mischief, and are no longer forces of production but forces of destruction.––Karl Marx and Friedrich Engels (1970b [1845-1846]:94)

The motivations for warfare are numerous. From the accounts discussed above,

it often involves retaliatory actions for offenses or even actions perceived as offenses.

Gunther (1927:266) found the main reason or rationale for an attack was the refusal to

pay blood money, while Barnett (1955:267-68), describing central Coast Salish groups,

said that a young warrior simply would just want to “try his newly acquired spirit

powers,” and he would then try to recruit others to join him. He also described how

individuals would often want to correct a perceived imbalance of suffering:

A common cause of war attacks everywhere was grief over the natural death of a child or other near relative. The psychology was oddly logical. Reasoning that it was unfair for him to suffer so acutely while others were without grief, the bereaved person encouraged a murdering and plundering expedition to relieve his suffering by imposing it on others (Barnett 1955:268).

While these attribute the causes to revenge, grief, or plain desire, other

informants described that stores were sought, such as “dried food” (Suttles 1950[10]:38).

Rarely was territory the primary motive. As Suttles (1951:321) described, “At least one

–– 127 ––

informant denied that people ever fought for territory, saying that they fought only ‘to

make themselves big.’” However, Suttles did record some fights, extermination, and

takeovers of territory. In one account, the Sooke arranged a marriage with the Makah in

order to fight the skwa'nǝ'nǝs:

After the whites came, the skwa'nǝ'nǝs were living at Sooke Bay and the Sooke at Sooke Harbor. A Neah Bay chief came to Sooke Harbor and the Sooke chief gave him his daughter in exchange for killing off the skwa'nǝ'nǝs. The Neah Bay chief said he would do so in four days. On the fourth day the Sooke were up early in the morning, listening. They heard the sound of shots coming from Sooke Bay. The Makah had come and cleaned the skwa'nǝ'nǝs out. After that the Klallam fought the Sooke. Klallam from Port Discovery and possibly elsewhere came over and attacked them. The Sooke chief named wa'nsiǝ escaped and walked through the mountains to the Songhees but the Klallam captured some of the Sooke and took them back as slaves and sold them to the south so that they reached the Columbia River (Suttles 1951:9).

The Klallam ended up gaining territory across the Strait of Juan de Fuca. As one

informant put it, “Becher Bay was all Sooke before Kl. [Klallam] came. But all settled

here at river. [B]ecause they fought for it” (Suttles 1952[12]:33).

The Lummi, according to one of their oral histories, acquired access to the

mainland through warfare as well. A warrior named Skalaxt, to avenge the death of his

brother, trained for years at a lake on Orcas Island, and returned to lead an attack on the

whole Skalakin village in which the killers of his brother lived. The Lummi killed most

of the people there, but were not satisfied, as they were aware that relatives of the

Skalakin lived in other villages (Stern 1934:119). He led yet one more damaging attack

upon another of their villages, and then a third time, he headed an expedition for a wife

among one of their villages on the Old Nooksack (or Red) River. The siem there offered

Skalaxt a wife in order to prevent a third such attack, and he accepted the marriage. In

Stern’s (1934:120) telling, “The siem told Skalaxt to take his daughter and accept the

river as a gift with her. He urged him also to move his relatives to the mainland and to

renew friendly relations with this tribe.”

So, in that account, the original or true motivations are multifaceted. Territory

was gained by the Lummi, but the initial reason for the attacks was accorded to revenge –– 128 ––

for the death of his brother, and gained territory was not the intention at all, but rather

was the side-effect of the successful outcome for Skalaxt––the expansion of territory

itself was a ”gift.” Similarly, the Sooke, whom the Klallam finished off, were attacked for

being so warlike, and they moved (or took over) Becher Bay since it was “deserted”

(Suttles 1951:9-10) or “unoccupied” (Gunther 1927:179). Curtis (1970 [1913]:22) noted

that the Klallam had harassed the Sooke for a long time as they had killed the eldest son

of a chief at "Chihwitsun" or Tse-whit-zen, attacking small parties of Sooke people when

out fishing or hunting; “In time, it is said, they accumulated on the beach just above high

tide a row of more than a hundred heads bleaching into white skulls.” Again, this is

another case of territory acquired, but with motivations that could be attributed to

revenge––the Sooke deserved the attacks, and the territory gained was a secondary

outcome.

These accounts indicate the multifaceted nature of reasons for engaging in

warfare. To invoke territorial acquisition or revenge as a single “cause” would not only

be simplistically reductionistic, but also inaccurate. A more encompassing way to assess

these motivations for gain is through the exchangeability of the various forms of capital.

The Various Forms of Capital

The nature of Coast Salish exchange incorporates various forms of capital (Figure

4). First, there is the natural capital available in Coast Salish territory; these are the

berries, salmon, deer, timber, and lithic resources, for example. These natural capital can

be converted into economic capital through household production: berries gathered and

dried; salmon caught, cut, and smoked or dried; the deer caught or killed, dressed, and

its meat smoked; timber cut down and processed into planks or a canoe; and lithic raw

material reduced to cores, bifaces, or specific tools. Natural capital is the potential

capital in the territory, or physical realm, but it takes productive activities to convert

natural capital into economic capital. Moreover, while some may be readily available––

as many bushes contain seemingly uncountable amounts of berries or a fishing station

may view a seemingly unending stream of fish moving upstream during peak seasons–––– 129 ––

ECONOMIC CAPITAL

CULTURALCAPITAL

SOCIALCAPITAL

NATURALCAPITAL

SYMBOLICCAPITAL

SPIRITUALCAPITAL

PHYSICAL REALM

SPIRITUAL REALM

Figure 4: Model of the exchangeability of various forms of capital for the Coast Salish.* *Arrows indicate potential flow or direction for capital; modified for this study after Bourdeiu (1977, 1990).

it often takes great productive capacity involving a whole household to convert much of

that natural capital into economic capital. The point is that a household may have lots of

natural capital available to them, in territories that they own or steward, but lack the

ability to effectively convert those resources into economic capital, particularly in short-

lived seasons when berries soon rot or dry up and salmon runs quickly pass through.

Natural capital, from the physical realm, must be converted into economic capital to

become part of the human cycles of exchange. Economic capital, here, refers to the

goods from production. It also can indicate the means of production; for example,

owning a reef-net, bird net, or a harpoon.

–– 130 ––

Also, for some, to be able to access natural capital requires that they know the

owners and have established relations with them. Marriages typically establish these

protocols, so that each family gains access to the other’s territories; this is social capital.

A group may have to expend some of their economic capital produced as a gift, or share,

to the owners for being able to access the natural capital in their territories. Cultural

capital indicates the knowledge that involves proper execution of any activity or

ceremony. Fishing may need to begin with a proper first salmon ceremony, for instance.

Or, simply, the activity requires specific knowledge or training for efficiency or success,

which could relate to technical skills, ecological knowledge, proper etiquette or

language, or magical phrases to be uttered or rituals to be followed. Cultural capital can

refer to the rights that a group or chief may have to a territory.

Symbolic capital indicates items are indicative of prestige and status, particularly

elite wares or items that have aspects that extend beyond functional or nonutilitarian.

Wealth, among the Coast Salish, has been associated with symbolically representing

another form of capital: spiritual. As natural capital is available physically, there is a

realm of spirits that also are broadly available metaphysically. However, instead of

productive activities, spiritual capital is sought through training and questing, or a spirit

may impinge upon a person, even unwillingly. Success or wealth, for the Coast Salish,

is an indicator of great spiritual capital. To become part of the cycle of exchange,

spiritual capital becomes symbolic capital: a longhouse dancer symbolically indicating

his spirit power in song; a hunter through a successful hunt; or a warrior through a

successful battle. Each of these instances represents a person with great spiritual capital

that enabled their success. That is, their spiritual capital is indicated by symbolic capital,

just as natural capital must be converted into economic capital to be exchanged with

other forms of capital. Natural capital is converted into economic capital through

productive activities, whereas spiritual capital is converted into symbolic capital

through conductive activities: individuals could attempt to attract spirit powers by

conducting themselves in proper manners, seeking solitude, fasting, and “training.”

–– 131 ––

Once acquired, the spirit power conducts through them, enabling their successes.

This process, wherein items of various forms of capital are exchangeable, for

instance, allows food items from household production to be “converted,” into wealth

items (Suttles 1987a [1960]:22) which in turn can be, through the potlatch, converted into

high status: “the relationship of food, wealth, and high status is complete. They all form

a single system.” Suttles further described the importance of affinal groups, or groups

allied through marriage:

The system of exchange and potlatching seems to have worked in this fashion. A kin or local group with a temporary surplus of any item of food, say the result of the run of some fish or the ripening of some berry, might take this surplus to an affinally related group in another locality and receive wealth in blankets or other imperishable items in return. Later the recipient in the first exchange might make a return visit with its own extra food and get back wealth for it. Any group that produced more food than its various affinally related neighbors would of course in time come to have more wealth. But any tendency for wealth to accumulate in a few places was controlled by the practice of potlatching, whereby the wealthy group converted its wealth into high status at the same time giving the other groups the means of continuing the process (Suttles 1987e [1960]:31).

Phrased in practice terms, economic capital (food), are exchanged with allied

affines (who make up the group’s social capital) or converted for wealth items (the

group’s symbolic capital). In a potlatch, these wealth items can be converted, by giving

them away, into other forms. Any of these types of capital indicate spiritual capital,

proof of the support or ability from their spirit power. Those with more cultural capital

and social capital are able to produce more wealth. It is their hereditary rights to

productive territories, a form of cultural capital, that allows them to produce more food,

the size of their household (also social capital) contribute to more economic capital. An

important aspect of this is the exchangeability of these forms of capital. As Bourdieu has

stated, “we see that symbolic capital, which in the form of the prestige and renown

attached to a family and a name is readily convertible back into economic capital, is

perhaps the most valuable form of accumulation in certain societies” (Bourdieu 1977:179;

emphasis original).

Bourdieu’s ideas about the exchangeability of capital are important for assessing

–– 132 ––

motivations in warfare. Warfare may be undertaken for plunder or slaves, but if the aim

is to exchange loot for a potlatch in which one gives it all away, the motivation would be

for status, or symbolic capital. As Mitchell (1984:39) pointed out, a “close relationship ...

existed between predatory warfare for the acquisition of slaves and the maintenance or

enhancement of high social status.” He argued that the main engine for raiding on the

Northwest Coast was for the “swift road to personal fortune” in slave-trading (Mitchell

1984:46). Combing through hundreds of ethnohistoric documents, Mitchell documented

the extent of this predatory warfare for the colonial period. All cultures from the Tlingit

to the Chinook engaged in the trade. Slaves were desired for their value as wealth, and

as symbols of wealth. Important slaves were worth more as captives that could be

ransomed at high prices. Mitchell effectively described a secondary system of

redistributive exchange. The potlatch serves as a redistributive exchange among allies

and kin, an internal network. Predatory warfare served as an external network, what

can be viewed as an exchange with one’s enemies, through force.

Mitchell (1981) reported that these are integrally related, as his account of the

Gitxaala (formerly Kitkatla) chief, Sebassa, revealed. From their base in the north coast,

he led a raid on the Nawitti on northern Vancouver Island, killing many men and taking

twenty women for slaves. These were later traded to Stikine Tlingit chiefs for furs,

which they used to buy commodities at Fort Simpson, and which Sebassa used for

potlatch gifts. Normally, preparations for a potlatch through productive means could

sometimes take years to amass the amount of wealth items. However, capturing slaves

through raiding and warfare provided a short-cut, creating immediate wealth.

Moreover, slaves were more than commodities, as they could be used to produce wealth

with their labor. With the conversion of slaves into potlatch commodities, Mitchell

demonstrated that warfare contributed to social capital, and thus was not simply part of

a vulgar materialist logic, but rather exhibited a social calculus as well. Since capital is

exchangeable from material to social and symbolic forms, even symbolic forms of capital

can be materialist in the economic sense.

–– 133 ––

Among the Coast Salish, similar economic exchanges occurred with the bounties

of war. Frank Allen noted that: “Killers like that č'uxe'lǝm [a Cowichan warrior] used to

get rich. They got slaves and goods of all kinds” (Elmendorf 1993:128). He aptly

described how warriors could gain capital and exchange it through a recounting of a

Klallam-Cowichan war that started near Port Townsend about 1850 (Elmendorf

1993:132-136). It provides not only a good example of exchangeability of economic

capital (or what may be better argued as social capital in this story) into other forms. It

is also illustrative of the protocols and cultural practices used as conflicts arise and are to

be resolved, so I provide some detail on how the conflict begins.

According to Allen, several Cowichan were present at the Klallam village for a

night of gambling, the bone-disk game of slaha'l. As the game continued, tensions

mounted––a fight erupted whereby several of the Cowichans were killed. Most of the

Klallams set about beheading the dead, but Klallam Pete, Xe'tanǝxw, an “awful bad

fighter,” had killed the son of a Cowichan chief. Instead of beheading him, Klallam Pete

kept his victim’s body, removing his entrails––he knew how to insert and apply certain

plants to keep it from decomposing for a time. By “playing” with it for many days in a

lake, he eventually acquired the dead man’s spirit power. The Cowichan made a

retaliatory attack, but called it off when the Klallam fired a hidden cannon that they had

borrowed from nearby settlers. Fearing what they thought must have indicated the

Klallams’ great warrior spirit power, the Cowichan turned their canoes back.

Eventually, the Cowichan chief sent some Skagit negotiators to get the body of his son.

Klallam Pete said “I want ten guns and one slave and one big canoe, and then [the chief,

c'o'sia] can take the body of his son in good shape” (Elmendorf 1993:135).

So after a while the interpreters come along again. Here comes a big canoe to Port Townsend. A big canoe now, with more people bringing the slave and those ten guns. When they bring that slave and that canoe and those ten guns, the Skagit chief comes along with them. He says, “Now, you Klallam people, that’s done fighting with the Cowichan people!” “Yes, that’s done.” This Skagit chief's name is kw'ałqe'dǝb. That’s the Skagit chief telling the Klallam no more fighting with the Cowichan.Duke York [the Klallam chief] says “All right, you kw'ałqe'dǝb, no more

–– 134 ––

fighting....” So that’s done now; so c’o’sia gets his son. No more fighting with the Cowichan now. Now xe’tanǝxw gets that slave and those guns. He gives the slave to Duke York and divided the ten guns to his people, one gun to each man. The canoe, that is all he kept for himself.Now, that is done (Frank Allen [Elmendorf 1993:136]).

Most of the fighters during that slaha'l brawl were satisfied with the symbolic

capital they gained from beheading their victim. Klallam Pete, having killed a chief’s

son, saw a greater opportunity for more capital to be gained. He gained not only

spiritual capital––through the macabre machinations needed to wrest the victim’s spirit

power from him––but also ransomed the corpse (as perhaps social capital, instead of

economic) for economic capital (guns, a slave, and canoe). He then gave away nearly all

of the guns, transforming them into prestige, or symbolic capital. All of this was a

verification of the power of his warrior spirit, or spiritual capital.

In assessing motivations for warfare, the focus should be on the ultimate

exchange of the goods acquired in warfare. If the initial or ostensible aim was for

plunder and/or territory, one could argue that the intention was for access to economic

or natural capital. However, if the warrior is using these goods gained from the raids,

not for their own individual consumption, but to exchange that capital through a

potlatch, then the ultimate exchange was for status and prestige, or symbolic capital. In

potlatch ceremonies, people gave away the wealth that was gained in exchange for

social and/or symbolic capital. In this economy, much of what was exchanged is

symbolic, even if the goods needed to acquire that symbolic capital are manifestly

material.

The potlatch presents a ceremonial exchange that allows for individual

expression of power while also contributing to the broader good. As Suttles (1987a

[1960]:23) noted:

Looking now at that most famous institution, the potlatch, I find that within this total socio-economic system, its most important function is to be found neither in the expression of the individual's drive for high status nor in the fulfillment of the society’s need for solidarity, neither in

–– 135 ––

competition nor in cooperation, but simply in the redistribution of wealth.

Suttles opted against vying for competition or cooperation in favor of a

redistribution-of-wealth argument, but he actually has demonstrated something even

broader: this “total institution” in the Northwest Coast serves both competition and

cooperation at the same time. It allows for people to gain status (competition) through

benefiting the whole (cooperation), with the whole being their household and allied

households, and perhaps the village. The exchangeability of various forms of capital is

what allows this to happen. Such an economic exchange is not egalitarian because it

showcases elaborate expressions of individual status and prestige. On the other hand, it

is not hierarchically centralized, not allowing for the rise of uniquely powerful chiefs––

they give away their capital, empowering others with economic capital of their own as

they exchange it for symbolic capital. As Suttles (1987e [1960]:24) pointed out, this

ceremonial exchange only “extend[s] the process farther.” Nor is this social organization

class-based, as Marx described, as these are factions of household chiefs and warriors

competing against each other for greater capital. A fitting description of this complex

social organization is provided, instead, by anarchism. It is a system that encourages

individual expression and local autonomy, but also encourages widespread networks of

cooperative ties. Moreover, those in such social organizations actively resist the

centralization or overly concentrated forms of domination or authority, a role in which

Coast Salish warfare appears to play a part. In the next chapter, I discuss how this

unique sociopolitical organization came to existence in Northwest Coast prehistory and

how the role of warfare changed throughout.

–– 136 ––

Chapter VI: An Archaeological History of Warfare in the Northwest Coast

To some, warfare was the predominant state of humanity in the past, before

“civilization.” These times are conceived, Snyder (2002) noted, as exhibiting the

“Hobbesian dangers of anarchy,” where presumably past lives are “nasty, brutish, and

shorte” without the Leviathan of the State to ensure peace. Yet, such descriptions for

much of human history as anarchic are correct, although not in the sense of the term

anarchy as chaos, with its suggestions of a free-for-all, but rather in the sense that people

lived without formal governments.

In the literature on international relations, warfare is described as occurring only

in anarchy. By anarchy, the term does not indicate a society without government but a

political situation in which an overarching power does not have (or has lost) dominion

over other groups (Wendt 1992, Snyder 2002:7). Warfare, accordingly, is a criterion

indicating autonomy, which is just as some in anthropology have emphasized

(Malinowski 1936:444; McCauley 1990:1). Wars erupt between states or cultural groups

that have no rule over another, or, with revolution, between those who challenge the

authority of dominant groups, indicating their assertions of autonomy. Warfare also

had particular aims that satisfied other needs. It is better seen as a practice, a strategic or

tactical opportunity of individuals pursuing power and capital.

The development of warfare on the Northwest Coast has traces that are

witnessed nearly back to its earliest periods, but warfare expanded to higher intensities

of devastation and increased scales of organization within the last two millennia. In this

chapter, I describe the development of Northwest Coast cultures since New World

colonization while highlighting the role and evidence of warfare. I will discuss changes

in social organization through time, which before contact was largely anarchic, in that

there were no formal governments. Instead, the social structures within and between

–– 137 ––

households consisted of the social arena in which there were assertions of power and

autonomy. In the following, I look at the archaeological history of the whole Northwest

Coast to provide a broader context for warfare in the Coast Salish region; a general

sequence is provided for the region (Table 2).41

Initial Colonialization & Archaic Period (ca. 11,000 to 3500 BP)

People have inhabited the southern coast of British Columbia since the glaciers

began to recede during the Quaternary Period over 10,000 years ago, arriving through

an ice-free corridor across the Bering Strait of Alaska. The first inhabitants are even

thought to have arrived thousands of years ago by water, traveling along the coast line,

according to one theory. Archaeologists have evidence for early coastal sites that

suggest a pattern of island migration from eastern Asia and down the Alaskan coast into

the waters of British Columbia (Fladmark 1979). Glaciers covered most of the landscape

on the northern part of the continent, enough to actually lower the mainland coast with

their weight, but islands are argued to have provided refuge along the route (Erlandson

et al. 2007). Early sites are documented in the north, in the Queen Charlotte Islands and

in the islands of Southeastern Alaska (Fedje and Christensen 1999; Dixon 2001, 2008), but

in southern British Columbia, the sites are located on the mainland coast and along the

lower Fraser River with a subsistence more reliant on terrestrial rather than marine

resources. For this reason, the earliest peoples are thought to have arrived by a land

route over the Bering Strait and around the Cordilleran ice sheet during the latest years

of glaciation, from 13,000 to 11,000 BP. Yet, coastal migrants could have moved upriver

as well.

In the Fraser River Canyon, the Milliken site documents the earliest occupation

in southern B.C., with a date of 9,000 years BP (Mitchell and Pokotylo 1996). Other

41. For summaries of warfare in North America, see Lambert (2002). For a global perspective, begin with Keeley (1996), Otterbein (2000), Gilchrist (2003), Arkush and Allen (2006), or Otto, Thrane, and Vandkilde (2006).

–– 138 ––

Table 2: General chronology for the Northwest Coast and Gulf of Georgia.*

Millenia BP GeneralNorthwest Coast

Chronology

Gulf of GeorgiaChronology

0POSTCONTACT Postcontact

1 LATE PACIFIC Late Period (Gulf of Georgia)

(transition)

Marpole2

MIDDLEPACIFIC Locarno

3 Beach

4

EARLY St. MungoPACIFIC

5

6

7 OldARCHAIC Cordilleran

Culture8

9

10

PALEOINDIAN11

12

*For more details on regional chronologies, see Borden (1970), Mitchell (1971, 1990), Carlson (1996; also Carlson andHobler 1993; Carlson and Dalla Bona 1996), Ames and Maschner (2001), and Matson and Coupland (1995).

–– 139 ––

important early sites include Glenrose farther downriver, and Bear Cove, located on

northeast Vancouver Island (Matson 1976; Carlson 1979). These early sites contained

large leaf-shaped knives and spear points along with large cobble tools. Since the

Milliken site was occupied in summer, it appears that fishing for salmon extends back to

its earliest inhabitants, even if it may not have been a primary focus (Mitchell and

Pokotylo 1996). Their camps were small and they foraged broadly, with extensive range

of the landscape, indicating a highly nomadic existence. They are argued to have lived

in small egalitarian groups that drew upon a wide range of environments, with

resources from land mammals outnumbering those from the sea.

Coast-wide, the archaeological record reveals a broadly similar material culture

pattern, with early sites exhibiting chipped stone tools (pebble tools, leaf-shaped bifaces,

and scrapers), from Ground Hog Bay (Ackerman 1996) and Namu (Carlson 1996) to

Milliken (Mitchell and Pokotylo 1996) and Dalles, Oregon (Cressman et al. 1960). Early

sites have also been discovered along the Oregon coast (Hall et al. 2004, 2005). These

have been argued to be descendant from the first migration of peoples (likely Amerind)

that is exhibited in the Nenana Complex of Alaska, a likely precursor to Clovis peoples

(Matson and Coupland 1995:57-58). The only manifestations of Clovis in the Northwest

Coast is at Wenatchee, although Wright (1996) argued for a possible fluted point

recovered at Coquitlam Lake, northeast of Vancouver. During this early period, there is

no evidence of conflict and warfare from burials, settlement patterns, or weaponry to

suggest antagonistic relations during the period of initial colonization. If present, the

conflicts would have been of such small organizational scale (in Wolf’s [1990] terms) that

it is unlikely that such evidence would be encountered.

During the Archaic Period, however, differences occur archaeologically between

the north and south. In the north, the North Coast Microblade Tradition (emphasizing

microblades, with less of a presence of chipped stone) and the Old Cordilleran Culture

(with pebble tools, leaf-shaped bifaces, and other chipped stone tools) in the central and

south coasts, as defined by Matson and Coupland (1995:68-81). Both are examples of

–– 140 ––

nomadic foragers operating in small, likely egalitarian bands. The north coast is almost

exclusively coastal in settings, emphasizing both marine and land mammal hunting,

while the southern coast exhibits a combination of coastal emphases, as at Bear Cove

(Carlson 1979) and more inland-oriented sites such as Glenrose (Matson 1976).

During this period, the atlatl was also used, which increased the effectiveness of

the spear by providing greater propulsion. The name for the tool derives from the

Nahuatl word for “spear thrower.” It allowed for an extension of the arm during the

throw, giving the thrust of the spear greater leverage than the unaided human arm

could provide. For use with an atlatl, spears were shortened and projectile points were

smaller. The earliest evidence for atlatls in the Northwest Coast comes from the Dalles

in Oregon ca. 8,000 to 9,000 BP (Kirk and Daugherty 2007:60) and from Namu, Period 2,

about 6500 to 5000 BP (Carlson 1996:100). Borden (1968) discussed an atlatl recovered in

Skagit territory, which he argued was likely Locarno Beach Phase in age; however,

Fladmark et al. (1987) radiocarbon dated it to ca. 1700 BP, or the Marpole Phase. This

atlatl exhibited anthropomorphic art, likely indicating spirit power connections. It

extended the thrower’s reach with its 41-cm length. Another main advantage of the

spear-thrower is the spear’s ultimate force of impact, the penetration of which can

cripple and immobilize game: “The impact of a spear results in either immediate death

or sufficient blood loss to weaken the animal relatively quickly. Spears, in other words,

impart a knock-down force and open significant wounds” (Hitchcock and Bleed

1997:355; see also Cattelain 1997, Yu 2006).

The first evidence of violence or conflict occurs during the Archaic Period. The

most prominent example is Kennewick Man, the contentious burial discovered in south-

central Washington that dated to 9400 BP. A portion of a long and fully serrated,

willow-shaped Cascade Point was found embedded in his pelvis. Its penetration was so

deep, that it was interpreted as having been a dart propelled by an atlatl (Chatters 2000).

The wound had mostly healed over, but the projectile point was visible through

computed-tomography (CT) scans.

–– 141 ––

Other examples of conflict in buried human remains, however, are rare, mostly

due to their limited number prior to 5000 BP from these areas. Other evidence, such as

artifacts, also do not explicitly suggest weaponry, although tools usually associated with

subsistence (projectile points associated with spears and darts at this time, as Kennewick

Man’s wound reveals) can be used as weapons. Settlement patterns also do not indicate

warfare, because of the nomadism of these mobile hunter-gatherers, with sites consisting

of camps or processing areas instead of architectural features. Maschner (1997)

suggested that warfare was present in this period––Keeley (1996:186) also found that

most hunter-gatherers engage in frequent (50%) or continuous (20%) warfare, although

this is less than the percentages for chiefdoms and states. Much of this conflict would

have involved melee combat, not a specialized industry producing weapons explicitly

for warfare or specifically defensive settlements, nor a division of labour allowing

warriors as specialized positions.

Warfare may also have been less common for both regions simply due to low

population numbers. If people encountered other groups less often, there would be

fewer opportunities or need for conflict. If aggressive groups were encountered, there

would have been other open territories to move into. Avoiding conflict and aggression

would have been an ready option during this time, both externally, between groups, and

within groups, as groups could opt to fission, and avoid internal conflict.

Consider such fissioning through the framework of power. The nature of social

organization and fluidity gave individuals and families in bands great flexibility. There

were options to choose with whom to ally or not. As the territory was sparsely

populated, there were always options to leave the group to pursue their lives elsewhere.

This can be seen to be a strategic advantage to those wanting to minimize concentrations

of power in one or few individuals, which as Blake and Clark (1999) note is “one of the

most effective means” for groups wanting to minimize the power of aggrandizers. They

portray this as an example of egalitarian ideology, but it also indicates a phenomenon

greater than an ideology of equality. I argue that this fissioning practice enacts a

–– 142 ––

principle of anarchic social organization, which actively resists the power or attempted

authority of others, aggrandizers in particular. That is, the action of fissioning is not a

statement about how people should relate to each other equally (in fact, as is the case

with many hunter-gatherers, they likely did not engage each other on equal terms), but

rather fissioning is an act that actually rejects social relations with aggrandizers––an

action that rejects another’s claim to authority and attempts to dominate and govern.

It is anarchic social organization, which has a “centrifugal logic,” as Pierre

Clastres (1987, 1994) has described. Such a centrifugal practice of dispersion counters

“centripetal” attempts to concentrate power in the hands of a few. This helps to “To

assure the permanence of dispersion, of the parceling out, of the atomization of groups,”

to maintain local autonomy (Clastres 1994:164). It is in this sense that fissioning is an

anarchic practice, one that happens to maintain control of an individual’s autonomous

power (i) rather than relinquish it to others (ii).

Because of such options, an individual may have great individual power (i), but

be severely limited in his or her ability to impose that power over others (ii). This

further affects a potential aggrandizer’s ability to attempt greater forms of power,

involving the ability of people to organize into larger groups (iii), and, in turn, the ability

to control or alter social settings (iv). The environmental conditions, with openness of

territory and the nomadic mobility to move to resource rich areas, aided individuals in

their personal power (i) by offering a range of possibilities to pursue. These types of

options and relations likely would have predominated during the Archaic Period,

lasting for over four thousand years (9,000 to 4,500 BP).

Early Pacific (4500 to 3500 BP)

Around 4500 BP, the populations of both the north and southern coasts

increased. According to a model by Croes and Hackenberger (1988), populations slowly

increased at a gradual rate over thousands of years. Matson and Coupland (1995:142)

described populations during the Archaic as “settling in,” in which foragers seasonally

–– 143 ––

ranged within localized regions. Distinct patterns began to emerge during the Charles

Period (4500 - 3500 BP) of the central Northwest Coast, with particular archaeological

signatures manifesting in various phases: Obsidian Culture in the Queen Charlotte

Strait, St. Mungo Phase in the lower Fraser Valley, Mayne Phase in the Gulf Islands, and

Eayem Phase in the Fraser Canyon.

In northern areas, the North Coast Microblade Tradition declined, with

microblades mostly absent by 4000 BP, with distinctive phases appearing in the Kitselas

Canyon, or Gitaus Phase; near Prince Rupert, Period III; and the Queen Charlotte

Islands, Graham Tradition (Borden 1975; Coupland 1988a, 1988b). Subsistence patterns

similar to the previous period are present though there is an obvious increase in the use

of upper tidal zone resources such as bay mussel, which appear as light to moderate

lenses throughout these coastal sites. Subsistence likely was still forager-based, though

perhaps closer to collector strategies, given the decreased range of mobility and

centuries of local ecological knowledge (Matson and Coupland 1995:114).

Warfare, during this period, was likely similar to the prior Archaic Period,

although one could argue for its increase due to the greater population and inter-group

contact. One measure of intergroup contact is apparent in trade routes, particularly of

obsidian. Whereas in prior periods, obsidian was generally found in sites in the local

region up to distances of about 160 km, during the Early Pacific––and even from the

terminal Archaic (ca. 6000 BP)––obsidian from numerous sources is found throughout

the coast. Moreover, the source types, including those from Mt. Edziza and others from

Oregon, overlap in their distribution areas at this time (Carlson 1994). Most

archaeologists treat exchange as an example of friendly relations, however, Keeley

(1996:126) stressed that “The fact that exchange and war can have precisely the same

results is often forgotten by archaeologists.” He is correct in that it is difficult to discern

whether the exotic goods are those of trade or from the plunder of war. Moreover, as he

further pointed out, “Contrary to the usual assumptions, exchange between societies is a

context favorable to conflict and is closely associated with it” (Keeley 1996:126; see also

–– 144 ––

Burch and Correll 1972). While it may not be possible to determine whether the trade

routes in obsidian indicate trade, warfare, or both, it does indicate increased intergroup

contact during this time.

It appears that some conflicts also did occur during this period. Cybulski (1994)

analyzed 57 burials that dated to this period and found that 21 percent exhibited

evidence of trauma likely due to interpersonal violence; for example, one male at Namu,

circa 4300 BP, had a bone point in his backbone (Hester 1978; cited in Carlson 1994). In

commenting on Cybulski’s analysis, Maschner (1997) noted that conflict appeared to be

associated with the central mainland, as sites on the Queen Charlotte Islands, such as

Blue Jackets Creek (Murray 1981), exhibited “little evidence” for such violence. Finally,

there are also some indications of weapons for war. Despite a lack of evidence of

trauma, warfare may be indicated at Blue Jackets Creek with two carved daggers that

accompanied a male burial (Severs 1974; cited in Carlson 1994:348).

The absence of microblades on the north coast is striking and does represent a

discontinuity in comparison to later periods. Microblades, for instance, have been

argued to be associated with Athapaskan groups, as the North Coast Microblade

Tradition (8600 to 4500 BP), which is seemingly derived from the second migration of

Athapaskan groups into the continent represented by the Denali Complex in Alaska

(Matson and Coupland 1995:82-84). Borden (1979:970) noted that “Later intruders, like

the Athapascans, greatly disrupted and complicated the distribution of ethnic groups.”

However, despite such a discontinuity, arguments have not been made for the

decimation of North Coast Microblade groups by others. Given the Athapaskan

presence of groups in the northern interior, it seems that their mode of subsistence

changed, perhaps more so through the increased interaction of peoples across the

coast.42

42. The use of microblades does continue, although in different contexts, as Matson and Magne (2007:142) suggested their association with later Athapaskan groups from 2000 BP. Also, microblades were adopted by Salishan groups during later periods, as recovered from Locarno Beach and Marpole sites, although not in great numbers (Burley 1980:21) and technologically derived from different microblade cores (Matson and Magne 2007:142).

–– 145 ––

With increasing population, the degree of social circumscription must have

added to a narrowing of options for those wanting to emigrate from potential conflict. If

people occupied adjacent areas, this scenario likely heightened the number of group

encounters and interactions and opportunities for potential conflict, with peoples likely

less related or at least less known. Remaining within one’s original territory may have

reduced potential conflict, even if it meant accepting or submitting to another’s claim to

power. As Blake and Clark (1999) described for Formative Mesoamerica regarding the

undermining of reigning egalitarian protocols, social circumscription in the Northwest

Coast appears to have allowed avenues for those seeking power to undermine not only

the egalitarian ideology but also to capitalize upon the limits of other’s autonomy. That

is, the range of options or practices available to those individuals and groups are

reduced; thus, their individual power (i) is subsequently reduced.

Having said this, I do not think population increases should be considered the

primary cause of warfare. As Keeley (1996:118-19) has observed, there is “some

relationship” between population and conflict, but the nature of that relationship is

“very complex or very weak or both”; moreover, he stated that: “In the broadest view,

the frequency of warfare and violence is simply not a consequence of human density or

crowding. However striking the images, human beings are neither rats packed in a cage

nor irascible billiard balls jostling on a table.” Moreover, such approaches ascribe the

cause to something external to humans, rather than as internal to human relations,

actions, and motivations. It is more productive to consider the environmental context as

a constraint or limitation that affects the options available to some and increases

opportunities for those able to apply their power over others. Changes in the

environment become opportunities for those to apply their tactics and strategies to their

advantage.43

43. An example from the Mississippian chiefdoms of the Southeast provides an illustration for such an approach. In the Savannah River Valley, Anderson, Stahle, and Cleaveland (1995) acknowledged that the Little Climatic Optimum was a favourable period for agriculture, resulting in surpluses of corn. However, using dendrochronological data, they demonstrated that even during that optimal period, droughts occurred––they were just less frequent than prior or later periods. In combining the chronology of chiefdom phases with the

–– 146 ––

Middle Pacific (3500 to 1500 BP)

Early Middle Pacific (3500 to 2000 BP)

At 3500 BP, significant changes occurred throughout both regions, which will

best be illustrated with examples from the Paul Mason Phase (3200 - 2700 BP) in the

north coast and the Locarno Beach Phase (3500 - 2400 BP) in the central coast. It is

during this period that significant economic changes are apparent, particularly for the

intensification of salmon. Salmon use has been noted in prior periods at sites such as

Chuck Lake in the north and Namu and in the central coast of B.C. Intensification

indicates a change in the nature of production that emphasizes storage.

Some arguments have been made indicating that the lack of salmon heads

indicates storage (e.g., Boehm 1973; Steifel 1985; Matson 1992; Matson and Coupland

1995:166); some have argued that site formation processes may affect the archaeological

presence of the frail bones of salmon heads, although at Crescent Beach heads were not

found in Locarno Beach Phase layers while they were present in the older St. Mungo

layers beneath (Matson, Pratt, and Rankin 1991). Matson (1992) advocated that salmon

intensification would also be associated with sedentism (minimally, to have a place for

storage) and, thus, houses or village would be present––and at Crescent Beach, a house

floor was uncovered.

On the north coast, the oldest known village is the Paul Mason site, located at a

prime fishing area in the narrowing of the Skeena River of Kitselas Canyon. Coupland

(1988a, 1988b) argued that the establishment of a village indicates a claim to that

dendrochronological record, they determined that chiefly elites were likely able to amass surplus stores to last three seasons of drought, although any extended periods would create havoc for their rule. Indeed, droughts in the dendrochronological data correlated with the cyclical rise and fall of chiefly elites in the Savannah River Valley, as indicated by various successions of short phases. Their case study and approach to the data indicates the important factor of environment, but they still locate the impetus for changes within the cultural context. Elite factions used such optimal periods to sustain their leadership and the failure of the economy often led to their downfall.

–– 147 ––

resource area. However, these claims were those of the group, specifically the

“egalitarian corporate group.” These groups, Coupland (1988a, 1988b) argued,

maintained an egalitarian social structure due to the similar sizes of houses, with a low

range of variance (58.6%).

In the central coast, the evidence for social ranking is apparent, in Fried’s (1967)

terms. Burials commonly exhibit labret wear, and the presence of elaborate zoomorphic

spoons at Musqueam NE and Pender Canal (ostensibly for feasting) indicate symbolic

differences within the population (Carlson and Hobler 1993). However, as Burley and

Knüsel (1989) argued, these status differences are achieved rather than ascribed. This

period exhibits a change in the mode of production to one that emphasizes surplus

accumulation. In the central coast, however, inequality is evidently more dramatically

emphasized. Such indications of status have occasionally been present in prior

periods––for instance, an elaborate grave for a young male at Namu (ca. 4000 BP; Curtin

1984). Regarding the increased sedentism in both areas, this period marks the

importance of ownership, as argued by Matson and Coupland (1995:152) for resources

that are “dense, predictable, and reliable.”

Ownership involves a claim to resources, claims that may be defended. During

periods of generalized foraging, the natural landscape had been open to foragers for

their use. With the construction of houses or other features on the landscape at or near

particular resource areas, labour had been invested into the landscape. These

constructions accomplish more than their functional appearance might indicate––these

were for more than subsistence or shelter. The invested labour also indicates a claim, a

symbolic mark to the landscape indicating that those who constructed those cultural

features occupy that landscape; this notion about the investment of labour as a claim to

ownership that can be extended philosophically back to Locke (1689).

Overall, the evidence for warfare remains thin during this period. There are

indications of Locarno Beach-like components at Shoemaker Bay and at Little Beach in

Ucluelet (McMillan 1982; 1996), which might represent the presence of Gulf of Georgia

–– 148 ––

Salishan peoples into the central and western coasts of Vancouver Island. Also,

Northern Wakashan peoples, about 2400 BP, expanded from the northern tip of

Vancouver Island into the Queen Charlotte Strait, and possibly northward, which

according to Mitchell (1990:357) possibly accounts for the Bella Coola isolate of the

Salishan speaking groups in the north. Such expansions, however, need not have been

of a violent nature. While there appears to be no evidence for warfare during this

period, the shift in economy to one emphasizing storage and surplus is a setting

conducive to warfare.

Earlier in this chapter, I described the development of more complex or unequal

forms of social organization with aggrandizers taking advantage of opportunities to

achieve dominance over others (ii). Environmental or social circumscription contributes

to a limitation of an individual’s power (i) and allows more opportunities for

aggrandizers to subvert egalitarian norms and dominate others (ii). However, simply

because there were elites, it does not necessarily indicate that others have been

dominated; rather, elite authority may have been earned. Chiefly elites can develop

their status through their ability to attract supporters and allies to their causes and

efforts. Contributing to the chief’s aims will eventually contribute to one’s own chances

for enhanced status. Such a logic is readily seen in the nature of how elites apportion

surplus to their supporters and household members. Chiefs can be indebted to their

supporters and allies, such that their status is purportedly higher, even while their real

authority is limited. Individuals may not be able to fission from those assuming power

over them (ii) as easily as prior periods due to social circumscription, but their options

for aligning with one chief instead of another remains open. Ethnographically, this

appears to be more the case for Coast Salish groups than northern groups, as they

reckoned kin bilaterally, allowing more options for individuals in their choice of whom

to align with, and which household to reside in. As Suttles (1987c [1960]:41-42) noted,

among Wakashan to the north, authority was established through the potlatch, creating

a series of ranked individuals. Farther to the north, among the Tsimshian, Tlingit, and

–– 149 ––

Haida, ranking was even more rigidly established through matrilineal descent, “making

alternative membership and individual mobility less possible” (Suttles 1987c

[1960]:41-42). During the Middle period, there was likely individual fluidity or mobility

for most peoples on the coast, at least in comparison to later periods.

With such mobility options still available, an individual’s power (i) and

autonomy is still strong, able to counter those attempting to dominate them (ii).

Therefore, chiefs would have had to spend more effort attracting potential supporters.

The more capital a person diverts (or redistributes) to his underlings and allies, the more

status he can retain. Supporters will follow a leader if they benefit from doing so (Clark

and Blake 1994:18-19). This status acquisition is a reflection of an individual’s

organizational power (iii), not necessarily one’s power over another person or group.

The power is also manifested in productive endeavors, such as constructing fish weirs,

houses, boats, or other activities.

This should not be equated with the view that elites are “necessary” for the

organization of such projects. For instance, some archaeologists have claimed, regarding

the Bronze Age in Europe, that leaders were merely managers facilitating the direction

of new construction projects such as irrigation canals and dams. Gilman (1981)

countered these functionalist notions of leadership, providing a Marxist analysis which

better explains inequality than approaches that argue for elites as functional and

necessary. Gilman (1981:3) noted “Even if one grants that certain economic situations

demand leadership for the common good, it does not follow that the rulers must be

recruited from a ruling class. It is not apparent that the best way of choosing efficient

managers is by birth.” There were elaborate burials for subadults and women,

something that elites-as-functional theories did not effectively explain. Moreover,

Gilman (1981) noted that it was not necessary for elites to spearhead large-scale

construction projects, as these could also be formed through the cooperation of

interested and mostly egalitarian groups, much as anarchists have advocated. However,

once these local groups invested their labour into such projects, Gilman pointed out,

–– 150 ––

they needed to protect their investments in order to continue reaping surplus from their

constructions. Neighboring groups might threaten to overtake the area (and use these

constructions without having to invest their own labour), or simply raid them to acquire

their increased surplus. Thus, a warrior class arose to answer such needs.

Similarly, in the Northwest Coast, the “functional” aspect of elites had been

promoted by Ames (1985). Such a model has been critiqued by Croes and Hackenberger

1988) and Coupland 1988b (also Matson and Coupland 1995:244-45), the former arguing

that large projects involving salmon intensification do not require hierarchical

organization. Instead these could be accomplished through self-organized groups, with

people acquiring status for their ability to organize (iii), not especially through a

particular function or direction in the project, but could be restricted through their

ability to convene others in egalitarian (or nonhierarchical) fashion towards a task, with

results to be redistributed accordingly.

I should note here that much of the anarchist literature explicitly concerns the

application of power, particularly a resistance to authoritarian power. Through such a

view, the Northwest Coast during the Early Middle period does not exhibit evidence for

authoritarian elites. Symbols for the identification of status are applied to adult

individuals that could have earned that status, rather than inherited it.44 For instance,

labrets appear to indicate items that identify Locarno Beach Phase elites, and these are

applied to perforations made in the lips. Percy (1974) determined that labrets were

worn by adult males; however, Cybulski (1991) noted that female burials with labrets or

labret wear were later found. Furthermore, Burley and Knüsel (1989:5) in the Gulf of

Georgia found that interment types during Locarno Beach were predominantly midden

44. Labret wearing generally consists of practice conducted and applied during the later stages of an individual’s life. It is interesting, however, that Duff (1952:80) had remarked that Chief Pierre of Hope, B.C., was said to have been “born with pierced ears and nasal septum”; that is, he was born already with these chiefly symbols, even though those are known by their very nature of the practice (i.e., perforation with cultural artifact) to be applied later in life––but in his case, in utero. Barring an interpretation of a miracle birth, the account appears to be an ideological move to justify a chief’s authority or eliteness backwards in time to make his authority naturalized, or more appropriately, supernaturalized.

–– 151 ––

burials with nearly 2 in 5 exhibiting grave goods (13/34; 38%), although those items

were “almost totally utilitarian,” likely indicating the nature of their skill in hunting,

carving, weaving or otherwise. Similarly, in the north coast, the houses of the Paul

Mason Phase (3200 - 2700 BP) are generally similar in size and there is a near absence of

prestige goods (Coupland 1988a, 1988b). The pattern generally extends to other

temporally related components at Haida Gwaii, Prince Rupert, and Southeast Alaska.

Late Middle Pacific (2000 to 1500 BP)

The developments that occurred during the Early Middle period established a

foundation for greater exhibitions of power and opportunities to undermine traditions

that favoured an equalization of power during the Late Middle Pacific period. Salmon

storage intensified, allowing for concentrations of economic capital. Moreover, in some

areas, notably the Coast Salish region, the prior centuries of intensified production set in

motion opportunities for elites to entrench their power, attempting to concretize the

authority they had earned or achieved and maintain it beyond even their lifetime––that

is, to make it hereditary. In the north, this period manifested as Period II in Prince

Rupert and the Kleanza Phase in Kitselas Canyon (2500/2000 - 1500 BP) and the

Marpole Phase (2400 - 1000 BP) in the central coast.

In the north, the McNichol Creek village in Prince Rupert area revealed evidence

for substantial inequality with one house exhibiting twice the size of the other houses; it

contains a significantly large central hearth, presumably for hosting elaborate feasts

(Coupland, Bissell, and King 1993). In a comparison of cache pit sizes between the Paul

Mason Phase and this period, Coupland (1998) demonstrated that the storage capacity

increased significantly from 38 m3 to 113.6 m3.

David Archer (1992) dated numerous villages in Tsimshian territory and found

that there was a common period of abandonment, between AD 1 and 400:

Of the village sites [in Prince Rupert Harbour area] that have consistent, reliable dates, there are 13 that were all abandoned within the first few centuries A.D. To place this in context, only one site was clearly

–– 152 ––

abandoned before this period. There appear to be two or three that were abandoned in the centuries after A.D. 500 and before the advent of Europeans, but certainly no more than this. The evidence points to a dramatic change in the local settlement pattern (Archer 1992:15; cited in Marsden 2001:100-101).

In an analysis of oral histories, or adawx and at.oow of the Tsimshian and Tlingit,

in comparison to the archaeological record, Marsden (2001) noted that warfare recorded

in their traditions corresponded to the village abandonments, reflecting temporary

control of the Prince Rupert Harbour by the Tlingit, as Tsimshian groups retreated up

the Skeena River. Later, according to the adawx, they are led by a warrior named Aksk

who reclaims their territory as Tsimshian (Marsden 2001).

In the central Northwest Coast, Croes and Hackenberger (1988) noted that the

Marpole Phase indicated perhaps the greatest period of salmon intensification in the

region, even when compared to later periods. Inequality at this time is represented in

burials by forms of cranial deformation (Mitchell 1971)––labrets cease by 2000 BP––and

there are burials with elaborate grave goods, even for women and subadults. Based on

this, Burley and Knüsel (1989) argued that substantial inequality and stratification was

indicated. During the Locarno Beach Phase––when status appeared achieved by the

symbolic use of labrets, for instance––the number of available status positions seemed

limited, though open to all. With status markers such as that of cranial deformation,

which can only be applied to infants, such markers of status would not be applicable to

everyone, suggesting that positions, resources, and the means of production were not

available to all. Cranial deformation is not present in the northern Northwest Coast,

although arguments for stratification can be made through other lines of evidence,

particularly in the form of elaborate graves for subadults. Notably, in this period, a

warrior complex appears to have emerged. This is best illustrated in the warrior’s burial

cache at Boardwalk dating to approximately 1800 BP, which contains a zoomorphic

stone club, a long basalt dagger, a “braining stone,” an orca jawbone club, and copper-

lined rods apparently for use as armour (MacDonald and Cybulski 2001:8-9). It is

perhaps to this period that the Hagwilget cache of stone war clubs belongs. Found in

–– 153 ––

1898 in Hagwilget canyon, these clubs exhibited elaborate zoomorphic imagery. Duff

(1975) aptly noted that, while these were clearly stone clubs, they were not quite

practical and more likely had symbolic functions: “One of the most curious things about

the stone clubs is that they are not so much functional weapons as they are images of

weapons.” Several of these “death bringers,” as Duff (1975:114) titled them in Images:

Stone: B.C., do appear unwieldy.

Beyond weaponry, or the imagery of war, the burial evidence suggests a

significant increase in warfare at this time. Cybulski (1994) noted that nearly a third

(32%) of adult males exhibited violent trauma in the north, with many burials exhibiting

dental fractures, forearm “parry” fractures, and skull indentations (which seem to match

the signature of a stone club). There are “trophy skulls” at the Boardwalk site

(MacDonald and Cybulski 2001), and three individuals at the Lachane site were

decapitated, dating to 1750 ± 40 BP, suggesting prisoners of war (Cybulski 1996).

Coupland (1989) argued that social inequality may have been been enforced

through coercion, indicating that these examples of trauma may represent internal

violence rather than warfare as elites may have sustained their dominance through

violent means. However, he also discussed how warrior ideologies may have been

directed outward in order to justify their position and minimize internal disputes.

Warfare appeared to be a strategy for leaders enforcing power, enabling them to

overcome egalitarian ideologies and strategies, which can be otherwise difficult to

undermine (Blake and Clark 1999). In the central Northwest Coast, however, the

evidence for such trauma is more limited, with only 10% of adult males indicating death

by violent trauma (Cybulski 1992). Instead, the social inequality of these cultures seems

to have been maintained more so through regional interaction including trade and

marriage, establishing networks of alliances participating in a regional ideology as

indicated by the prevalent Marpole art style, for instance, in stone bowls and burial

patterns. Both Grier (2003) and Brown (1996) argued that the inception of regional

interaction networks that Suttles described for the ethnographic period appeared to be

–– 154 ––

extant during the Marpole Phase.

The iconography in the art seems more representative of spirit powers (e.g.,

Carlson 1983) than it does for imagery of warfare, although particular spirit powers have

been important for warriors among the Coast Salish, as discussed above. Perhaps this

regional network of alliances staved off military expansion. Mitchell (1990) argued that

during this period (ca. 2400 BP) there was an expansion of Wakashan peoples from the

northwest coast of Vancouver Island into the Queen Charlotte Strait. There is a

significant discontinuity between the Obsidian Culture type emphasis on chipped stone

to the Queen Charlotte Strait type (with the near absence of chipped stone; bone and

mussel shell tools); McMillan (1999) and Coupland (1989) both argued similarly for a

Wakashan expansion at this time to the Washington coast, as indicated by the presence

of Makah populations. However, there is a lack of evidence for warfare in the Gulf of

Georgia (e.g., burial remains, artifacts, or settlement styles); while in the north, a

“warrior complex” is clearly present. This led Coupland (1989) to argue that warfare

rose in tandem with the development of social inequality in the north, but apparently in

the south, warfare appears to follow the development of social inequality. It is during

the next period, the Gulf of Georgia Phase in the central coast and Period I in the north,

that the evidence for warfare is most apparent, particularly in the construction of

defensive sites.

On the development of an elite class

It has been argued that the power gained from the intensification of storage

practices and the accumulation of economic capital led to the establishment of an elite

class, particularly in the Gulf of Georgia. Symbols of eliteness are applied to subadults

as indicated in the burials. In fact, the symbolism of cranial deformation was applied to

infants. As Burley (1989) argued, these indications reveal ascribed status. The symbolic

practice that may have been applied during an individual’s lifetime, the labret, is even

abandoned. This suggests, à la Fried (1967), that the availability for elite positions had

–– 155 ––

been narrowed significantly. In an extreme sense, no longer could one earn the status

shown by a labret or other elite symbolic items; instead, one had to be born with it.

After Mitchell (1971:54), Burley and Knüsel (1989) provided a solid argument for the

differences between achieved and ascribed status differences in Locarno Beach and

Marpole, one that was supported and advanced by Matson and Coupland (1995:215).

According to Roscoe (1993), who advocated a primary model for practice theory

in anthropology, the political concentration of power involves three predominant

variables: Effectiveness, or the degree of power; extent, the range of the applicability of

power; and its institutionalization, establishing the rules of authority into structures

external to the personal characteristics or achievements of the powerful individual into

laws or official roles. However, initially aggrandizers for power have to continually

create and maintain alliances, which are costly, and some end up owing their supporters

more than they have credit for––an unstable form of power. It is more effective,

therefore, to augment power by institutionalizing dominance beyond the charisma and

authority that one has achieved:

Practice dictates that leaders will take advantage of whatever opportunities exist to replace direct control of others with control of structures that dominate them, because thereby their dominance becomes more efficient and their interests are advanced (Roscoe 1993:118).

In the Marpole Phase, we have evidence for the development of all three aspects

of political concentration: Effectiveness is evident in the widespread appearance of elite

wares; extent is apparent in the networks of elite exchange, which through regional

alliances of elites the range of such power expands; and, finally, institutionalization is

suggested by the establishment of hereditary eliteness in the symbolic practice of cranial

modification and subadults with elaborate grave goods. Just as Clark and Blake (1994)

argued that elites in coastal Chiapas overturned egalitarian levelling mechanisms to

allow for the rise of aggrandizers, Marpole elites undermined the mechanisms that

maintained status as achieved, allowing for the stratification of a hereditary elite. In

institutionalizing their authority beyond the abilities or charisma of an individual, they

–– 156 ––

found a less costly way of maintaining control: by hoarding symbolic capital in the

nature of the elite wares (e.g., stone bowls) and practices (e.g., cranial modification),

enacting hegemonic controls that constrained the participation of non-elites.

Interpreted through Wolf’s (1990) model of power, Locarno Beach Phase elites

appeared to have organized their power (iii) through their achievements, their own

personal ability (i) and building upon their dominance over others (ii). In such a system,

they needed to maintain their personal power (i), their dominance over others (ii), and

maintain organization through alliances and gift-giving (iii) feasts, a process that

demanded much effort and continual renegotiation, particularly in a society where one

gives away his or her economic (i.e., material) capital in exchange for social, cultural,

and symbolic forms. During Marpole, accordingly, the networks of exchange must

continue, however, the need to display individual power (i) is lessened––it does not

need to be achieved through years of experience demonstrating an aggrandizer’s power

over others (ii); rather, it has been institutionalized or hegemonically ritualized through

the hereditary inheritance of that individual power (i). The elite infant is born with that

power. This ideology of ranking maintains that the infant already has some power over

others (ii), that is, over non-elites. Marpole elites organized their power (iii) in such a

manner as to control the settings for participation in that eliteness––in other words, they

applied structural power (iv).

On storage and its implications

Much has been made about the onset of storage on the Northwest Coast and its

implications that storage allows for higher population numbers (e.g., Croes and

Hackenberger 1988) or increased carrying capacity, just as it has been known to have

implications worldwide in the past (Kelly 2000, Otterbein 2004), as discussed above.

Population numbers, while increasing the episodes for inter-group contact, does suggest

that the chances for tensions are increased. However, increased population does not

necessarily mean that such contacts will be violent (Keeley 1996:118-19). It is perhaps

–– 157 ––

more accurate to discuss how storage contributes to the abilities of those aggrandizers

aiming for more power, and to view increased population associated with storage as

altering the conditions and dynamics of the social field. The importance of storage does

represent a key moment, with its implications for the accumulation of capital.

Some archaeologists like Coupland (1988a, 1988b) and Matson (1992) have

argued for the significance of the storage of salmon as similar in its effects to the onset of

agriculture, in line with Schalk (1977). Accordingly, it is not agriculture per se that is

important but rather storage of surplus which is critical, providing a basis for social

complexity; factors necessary for storage included dense and predictable resources.

Consider the introduction of storage through the frame of power and practice. The

storage of material foodstuffs is the amassing of economic capital. This is capital, again,

that can be converted or exchanged into other forms of capital. One could feast another

chief, establishing an ally (social capital) or trade for knowledge or advice (cultural

capital). Moreover, the practice or process of feasting and gifting itself––in public

display, meaning in view of witnesses––increases an individual’s status (symbolic

capital).

Storage allows for greater display, or gifting episodes, something which occurred

earlier Mayne Phase/Locarno Beach, for instance, at Pender Canal, where Carlson and

Hobler (1993) made a case for the “feast[ing] of the dead,” from the presence of

elaborately carved mountain goat-horn spoons. However, most funerals, while

ostensibly about the dead, are more accurately for the living. Similarly, these spoons

were more likely indicated feastings by and for the living. Storage is a practice that

enables the power of an individual to increase, allowing for public displays (which, in

themselves, are demonstrations of organizational power [iii]). Among the the Coast

Salish, as other Northwest Coast groups, this witnessing concept is critical. It allows one

to publicly receive validation for the title one has gained, or rights acquired.

While storage has been important for the increase of social complexity, it also has

implications for warfare. Earlier I described warfare in relation to a mode of production

–– 158 ––

as opposed to a mode of destruction. The latter phrase, used by Ehrenreich (1997), does

convey the damaging aspects of warlike activities. However, it is only accurate in a

limited sense. To the victim of a successful attack, the results are destructive, however,

to the victors, the results are productive, although not in the traditional sense, as the

goods were not produced––at least not by the raiders. Instead, the raiders have decided

not to pursue a mode of production but a mode of acquisition, in directly acquiring

produced foodstuffs, or economic capital.

Late Pacific (1500 BP to contact)

The Late Pacific Period, beginning around 1500 BP, is marked by the widespread

presence of defensive sites across the Northwest Coast. According to Moss and

Erlandson (1992), around 1200 to 1600 BP, forts begin to be constructed in southeast

Alaska, for instance, but most were constructed from 600 - 900 BP. These dates match

those on the west coast, as documented by McMillan (1992) at T’ukw’aa refuge (870 - 100

BP); a site where the adjacent village predates the refuge use by centuries (1240 - 580 BP),

although he noted that the selection of such a site with a nearby refuge seemed to be in

their selection criteria. The site of Ozette is adjacent to Cannonball Island, which has the

oldest date (2010 BP) for a refuge site on the coast although this date may be related to

its function as a whaling lookout instead of warfare at this time (Moss and Erlandson

1992). Keddie (1984, 1995, 1996) dated numerous forts in southern Vancouver Island

that ranged 450 to 1160 BP. Undated rock-wall fortifications also likely date to this

period in the Fraser Canyon (Schaepe 2000, 2001, 2006). The introduction of the bow

and arrow, in conjunction with defensive sites, indicates a new form of warfare,

according to Maschner (1997), changing from mostly hand-to-hand combat to more

effective long-distance weapons (bow) and siege tactics, as indicated in defensive sites.

Throughout the world, the bow and arrow was widely adopted and even displaced the

atlatl and dart in many areas. In the Northwest Coast, use of both appears concurrent,

but the bow has been widely in use since about 1500 BP. There are many technological –– 159 ––

advantages. Although the bow is predominantly used within the same range as the

spearthrower (e.g., 20 to 30 m), the bow is a more accurate weapon (Cattelain 1997:230;

Yu 2006). The energy needed for propelling the spear or dart momentum is with the

throw itself, which requires space for the overhead throw; the process of the throw

inhibits concealment and the significant movement can alert prey. Less room is needed

to shoot the arrow than to throw a spear. In fact, a bowhunter can remain quite still

while releasing the arrow, since most of the propulsion energy is contained within the

bow and its string, when drawn. Shooting an arrow can be done much more quietly

than throwing a spear. The launch time required is significantly reduced with the bow

and arrow, which means that it can be shot with “far more secrecy than darts from an

atlatl” (Nassaney and Pyle 1999:259).

The increase in projectile amount arguably is the most important aspect of the

atlatl to bow and arrow transition. Arrows are light enough that they sometimes do not

alert prey (or scare them off) during the shot, permitting additional shots at the prey.

The smaller projectile and higher accuracy significantly expands one’s range of hunting

possibilities as well, since a hunter can target much smaller animals.

Wars on the Northwest Coast were fought with many types of weapons in the

precontact past: whalebone clubs, spears, knives, and the bow and arrow. Prior to the

introduction of the bow and arrow, however, battles generally involved hand-to-hand or

melee combat. Though spears were commonplace, they were used predominantly for

thrusting and stabbing––that is, in close range. If a warrior threw his spear, he also

threw away his primary weapon. According to Gunther (1972:14), the long flint-pointed

lances used in warfare on the Northwest Coast were used as spears, but were not

thrown. Similarly, Barnett (1955:270) described that “Spears were seldom hurled but

were used for thrusting,” while bows and arrows were used when the fighters were not

at close range. Slings were also used, and likely knives thrown as well, but

predominantly consisted of hand-to-hand combat. With use of the bow and arrow, a

dimension of combat would have come into play, as melee weapons sparred with

–– 160 ––

projectile weapons.

In other parts of the continent, it appears that warfare also increased with the

widespread adoption of the bow and arrow. In the Southeast, the Middle Woodland

Period ended as the Late Woodland began, marked by small, corner- and side-notched

points. According to some studies, there was an increase in mortality rates, particularly

among young men, during this time; indeed, many were found with projectiles

embedded in their bones at time of burial. It also has been argued that the palisaded

villages of the Southeast at that time were a response to the shift from melee combat to

projectile-based combat (Dye 2006). In an overview of North American data, Nassaney

and Pyle (1999:260) determined that the expansion of the bow and arrow likely indicated

a “qualitative reorganization in the scale and practice of warfare and/or hunting game.”

Warfare was also indicated in burial remains during this period. According to

Cybulski (1994), after AD 500, over a quarter (27.6%) of burials across the Northwest

Coast exhibited trauma in the skeletal remains, the highest percentage in comparison to

prior periods, although the sample size was smaller due to the shift to above-ground

forms of burial.

Warfare may manifest in more than weaponry and defensive sites. In an analysis

of sites in Kuiu Island and Tebenkof Bay in southeast Alaska, Maschner (1997) noted

that Early Period sites were located along the central portion of bays, close to the clam

beds. Sites during the Late Period, however, were located along spits and extensions

along the sides of such bays. These were not defensive sites per se, but their location

provided more visibility to the open seas: three times as much in comparison to

settlements deep within the bay, although inconveniently located away from main clam

areas. In fact, these sites were also more exposed to winds and storm waves. Maschner

(1996:187) described these changes in the Pacific III, or Late Period, as a “transition from

an economically maximizing settlement pattern to a pattern of political maximization.”

No longer were settlements based primarily upon ready access to resources, but on the

ability to defend those resources––the new settlement locations were the response to the

–– 161 ––

sociopolitical as well as the natural environment, or what he referred to as the “politics

of settlement choice” (Maschner 1996; also 1997).

In describing the settlement patterns of Gwaii Haanas, Haida Gwaii, Acheson

(2005:320) noted that “Personal protection was second only to access to resources in

determining settlement location.” Defensive sites, which were generally attributed to

the Late Period, occupied “small, precipitous islands, stacks, or headlands,” and these

sites were predominantly along exposed coastlines (Acheson 2005:321).

Around the vicinity of Lake Kitwancool, Prince (2004) noted the locations of

cache pits in remote and steep locales, which he argued were used for the protection of

foodstuffs from raiding during this period. The lake also contained a lookout site with

nearly full views of the lake’s perimeter. He noted that the Little Ice Age may have

contributed to regional tensions with neoglacial advances occurring in the larger region.

Maschner (1997) also noticed changes in subsistence patterns. In prior periods,

open sea-mammal hunting and fishing were emphasized, while in the Late Period,

riverine salmon fishing and land mammal hunting were emphasized. These indicated to

Maschner (1997) a shift to safer subsistence strategies, emphasizing collective labour in

safer riverine areas and in forests––much less riskier than individual or small group

activities on the open sea.

In the central coast, Pegg (2000) has argued similarly that warfare affected

gathering practices such as cedar-bark gathering. Using a wide database of

dendrochronological dates from culturally modified trees (CMTs), he found that there

were changes in the pattern and range of their use (which predominantly were

postcontact). Initially, he hypothesized that it correlated with disease, although he

found this not to be the case. Rather, Pegg (2000) determined that ranges for cedar-bark

gathering were more restricted during flare-ups of warfare after contact.

Other routine practices could also be affected by warfare. MacDonald (1979,

1989) conducted the initial excavations at Fort Kitwanga, a protohistoric fort in

Tsimshian territory. The fort was located on a high hill-top near the mouth of

–– 162 ––

Kitwankul River and consisted of at least five houses within a palisade. The team also

gathered oral histories and other information through ethnographic interviews from the

village of Kitwanga two miles downstream. From these informants, they were able to

interpret the use of such features as small sweat lodges and puberty huts. Normally

puberty huts would be located at some distance from the village, however, in times of

war the Kitwanga elders noted that the huts would be adjacent to the village, as found at

Fort Kitwanga. Thus, arguments for warfare during this Late Period can be extended

well beyond burial evidence, weaponry, and fortifications.

Territories were also expanded during this period. About 1000 BP, according to

Kinkade and Powell (1976), the Wakashan Makah expanded southward from Vancouver

Island to the Olympic Peninsula, overtaking Chimakuan areas. They used several lines

of evidence, including linguistics and oral histories to document Chimakuan place-

names and migrations among ethnographically Salishan Quileute and Wakashan Makah

areas. This suggests that whaling practices would also have been brought southward.

However, as Wessen (1990) noted, there is a continuity in whaling traditions on the

Washington coast for at least 2000 years.

In his prehistory of West Coast, McMillan (1999; also 2003) suggested that the

oral histories do tend to support a southward expansion about 1000 years ago or so.

This happens to correspond to the early constructions of defensive sites in Barkley

Sound (1200 BP) and on steep-walled Tatoosh Island, off the northwestern tip of the

Olympic Peninsula, which was inhabited about a thousand years ago (McMillan

1999:151-152). Elmendorf (1990:440) noted that the Makah and Klallam appeared

“intrusive” in the region, finding that both groups appeared to come southward from

Vancouver Island disrupting an apparent continuous region of Chimakuan peoples.

Postcontact Period

By the postcontact period, warfare is noted as a fact of life in the Northwest

Coast. Comparable to the introduction of the bow and arrow is the introduction of –– 163 ––

firearms to the region in the 1790s. Elsewhere, it is known that firearms caused

significant impacts. In the Gold Coast of Africa, firearms enabled the Asante expansion

to gather slaves for the trade (Wolf 1982). In North America, firearms were differentially

distributed to groups based on their allegiance to, or willingness to trade with, fur

traders or colonist groups (Worcester and Schilz 1984). In the Northwest Coast, the role

of firearms has been played down (e.g., Fisher 1976, 1977; Cole and Darling 1990),

suggesting that a flintlock musket would have been less effective within the damp

climate, noting that the musket was inclined to misfire. Cole and Darling (1990:126)

found that “for speed of fire, accuracy, unobtrusiveness, and dependability, the Indian

bow often had the edge as an offensive weapon.” They noted that on the West Coast of

Vancouver Island, both the Boston, in 1803 and the Tonquin, in 1811, were conquered by

the Nuu-chah-nulth with traditional weapons (Cole and Darling 1990:127).

Undoubtedly, Northwest Coast groups were proficient with weapons such as the

bow and arrow, however, this does not explain the widespread adoption of firearms by

these groups or the hesitancy in selling such weapons by traders, as at Fort Langley

(Angelbeck 2007:269-272). Fisher (1977) noted the disadvantages of the weapons, but

given their quick and widespread adoption, the advantages of muskets to native groups

appear much greater than Fisher allows. Moreover, other researchers have been clear

about the advantages of firearms in the Northwest Coast.

McMillan (1999:192-193) discussed how the early acquisition of firearms gave the

Tla-o-qui-aht an advantage over groups in Barkley Sound that “did not yet have

muskets.” The wars led to a devastating drop in the numbers of Toquaht, who had

“become few,” according to one of Sapir’s informants (McMillan 1999:193). In the Coast

Salish area, Curtis (1970 [1913]:20) described how the Lekwiltok, instead of attacking at

night by surprise, now attacked during daylight. Similarly, Taylor and Duff (1956) also

emphasized the advantage of firearms in the southern expansion of the Lekwiltok.

While the specific reasons for the onset of warfare during the post-contact period,

may relate to numerous reasons, it is clear that firearms significantly altered normal

–– 164 ––

arenas of interaction. Duff (1964:61) argued that the impact of firearms made intertribal

wars “much more lethal affairs,” finding that the mortality rates became much higher.

It is difficult to gain an appreciation of the destructiveness of this warfare without going over, one by one, the traditional histories of each of the tribes. Murders, massacres to avenge them, and more massacres in retaliation form a constantly recurring pattern. Many small tribes were, in effect, exterminated. Some of the more powerful tribes, or alliances of tribes, embarked on contests of mutual annihilation. The wars continued without abatement into the 1860s. In the early journals we find frequent comments about the constant fighting among the Indians, but these somehow fail to convey the extent of the slaughter that was occurring just beyond the gaze of the men in the trading posts (Duff 1964:61).

Duff was acknowledging how firearms affected traditional practices, producing

substantial effects in comparison to traditional forms of combat. Perhaps the desire for

retribution was just as great in pre-firearm days. However, there were equal types of

arms available to attackers and defenders––all groups could make spears, bows and

arrows, and knives: the means of destruction were available to all. Technologically, the

arena of battle was equal. Hence, there was a need to gain advantage by other means,

by surprise, with attacks occurring in the middle of the night or early morning. Even

then, many groups deployed watchmen, messengers, and scouts as a caution against

such measures.

With the onset of firearms, the distribution––as occurred throughout much of

North America since colonization––manifested with certain inequities, and the effects

were substantial (Worcester and Schilz 1984). Collins (1950:337) stressed that the “The

role played by the introduction of the gun should not be minimized in this new

emphasis on warfare.” It increased as northern groups were attracted to trade at the

forts, such as Victoria, bringing groups more commonly into interaction than before.

Warfare also increased among Coast Salish villages and “Slave-raids became so

prevalent that up-river peoples were afraid to make the trips to the salt-water sites [in

Puget Sound] which had always been part of their annual subsistence quest” (Collins

1950:337).

Warfare may have been fueled by individuals aiming to increase stores of

–– 165 ––

surplus, particularly for potlatches (Mitchell 1984). For example, Mitchell’s (1981)

account of Sebassa’s slave raiding shows how slaves were acquired and traded to other

groups for furs that could be traded at the forts for items to distribute at potlatches.

Warfare and slavery seem to go hand in hand, as slaves (noted by Donald 1997) were

always acquired through warfare (or traded for with those who did capture them).

Furthermore, Donald demonstrated that slaves were not simply symbolic of status, but

rather contributed significantly to the labour pool of those warrior elites. Leaders,

particularly aggrandizing ones, aimed to maintain their high status and increase power

by generating more surplus, which they used for alliance building and rewarding

benefactors. However, most chiefs in the Northwest Coast had only their households

composed of kin, and kin generally could not be forced to generate surplus at extreme

levels, as they have a degree of autonomy (i). However, with slavery, such bounds are

removed, and slave-owners could use the labour of slaves to productively increase their

surpluses for potlatching (ii). Surely, these are generating material capital for such

endeavors, but through gifting, these leaders exchanged material capital for symbolic

capital. For instance, Martindale (2003) argued that Legaic of the Tsimshian became the

leader of a paramount chiefdom, not through conquest, but through the cultural practice

of the potlatch. Through militaristic control of the access to the fur trade, he was able to

generate surpluses such that he could sponsor four sequential potlatches that inarguably

set him as the highest ranking chief among those groups for a brief time during the

colonial period, ca. AD 1825-40. Warfare it seems is used as an external means both to

acquire resources to maintain high status, but also to direct internal strife outwards

(Coupland 1988a, 1988b); this is an argument similar to that raised by Coser (1956) and

Otterbein (2004), mentioned earlier (see page 44), where external tensions increase

internal cohesion. With the manufacture of specialized clubs and the construction of

defensive sites, such projects also became another avenue for the allocation of labour,

opportunities noted for aggrandizers to assert their power (Arnold 1993).

Warfare represents a multitude of practices, tactics and strategies, that leaders

–– 166 ––

can use––not only for defeating enemies and acquiring loot––but also for maintaining or

building power within their societies.

Conclusion

As this archaeological history recounts, the range of practices open to individuals

changed through time. Of course, individuals could create new cultural practices as

well, and these are reflected most readily in the archaeological record through the

introduction of new technologies and the range of variability among certain artifact

types. But, as always the case, most creativity was not sui generis, but rather it was a

response, a synthetic resolution of the dialectical interplay between current needs and

past traditions. Most innovations are commonly said, to be done “on the shoulders of

giants,” a statement that recognizes that such creativity comes from incorporating

existing cultural practices or ideas. Besides, one works with what one has access to

(economic capital) or what one knows (cultural capital)––societal tradition influences

our formation as individuals––what Bourdieu (1977, 1990) referred to as habitus.

Bourdieu was emphatic that while everyone has a habitus, each still improvises among

cultural practices and ideas at hand. This concept comes from his structural influence in

Levi-Strauss, who introduced the concept of the bricoleur. Although Levi-Strauss

(1966:16-17) had primarily used it to discuss the creation of myths, as when he had

written that “Mythical thought is therefore a kind of intellectual ‘bricolage’.” Such an

innovator is the bricoleur, a handyman or craftsman that uses whatever is at his disposal

towards the project at hand––in Levi-Strauss’ case, primarily terms and symbols.

Bourdieu extended (or returned) the notion of the “handyman” to its materialist basis in

practice.

–– 167 ––

Chapter VII: Lookouts, Refuges, Fortifications, and Stockades

The range of Coast Salish defensive practices

In this discussion of defensive sites in Coast Salish territory I demonstrate that

while there are certain commonalities, there are also distinctions and variations, which is

similar to the distributions of other Coast Salish traditions and practices that I have

previously discussed (see Chapter V). Before I describe these defensive features, I

explain how archaeological features reveal defensive aspects, as archaeologists often

have questioned whether many sites, interpreted as fortifications, were associated with

warfare at all.

The Case for Defensiveness

While the defensive use of many archaeological fortification sites may appear

obvious, this interpretation is a matter of perspective. Several archaeologists have

recently published works challenging the omission of warfare from archaeological

interpretation worldwide. For instance, Guilaine and Zammit (2005) noted that in

Europe archaeologists have sometimes interpreted walled fortresses merely as well-

fenced farms, with turrets classified as granaries. Keeley (1996) challenged such

interpretations as well, considering much of the anthropological discourse to have

Rousseauian overtones. In his work, War Before Civilization, Keeley presented evidence

for warfare in the past. He found that sometimes archaeologists obscured the evidence

of war: weapons interpreted as ceremonial items; warrior graves as merely status

symbols; and even Late Neolithic battles axes “considered a form of money.” In one

example, he commented that one 5,000-year-old burial “was found with one of these

moneys mischievously hafted as an axe. He also had with him a dagger, a bow, and

some arrows; presumably these were his small change” (Keeley 1996:19-20). In the

–– 168 ––

American Southwest, researchers have disputed the popular notion of Pueblan peoples

as simply peaceful. Important work by LeBlanc (1999, also Rice and LeBlanc 2001) has

determined that much of Anasazi material culture is infused with contexts of warfare,

with defensive sites being constructed from ca. 1000 BC to about the time of contact.

Even the apparently peaceful period, the zenith of Chaco culture, was maintained,

according to Lekson (2002:611, 614), by a “socialization of fear” through violence, the

“extreme processing” of corpses, and possibly ritual cannibalism.

In the Northwest Coast as well, it has been necessary to challenge such oddly

pacifist interpretations of weapons. As discussed above, Fisher (1976, 1977:16-17)

argued that firearms were of little advantage over the bow and arrow and so had to

account for why the technology was desired. He offered that, rather than really serving

as weapons, these firearms likely contained “emotional value” or served as “phallic

symbols.” Archaeologists in the Coast Salish area also have argued against any

defensive aspects of trench-embankment sites. For instance, despite providing a fruitful

presentation on trench-embankment sites, Buxton (1969) ultimately proclaimed that

these were not defensive at all. Instead she argued that these trench-embankment

features were used for fish drying or for game drives. She concluded this despite

several lines of evidence to the contrary. For these subsistence-based interpretations,

several factors are not taken into account: (1) the sites often were located upon

landforms that were difficult to access; (2) these sites are not generally associated with

prime hunting or fishing areas (in fact, many were situated high above beaches away

from fishing or clamming areas) and usually distant from fresh water, which is not good

for hunting; (3) these sites often exhibit evidence of palisade walls at their perimeters; (4)

their middens have a similar diversity of artifacts and faunal material to residential sites,

if less substantial in volume (as opposed to a narrow range associated with a singular

subsistence activity); (5) and, lastly, there is a wealth of ethnographic, oral history, and

ethnohistoric information about the defensive use of similar features, while there is

limited evidence of such massive trench features associated with fish-drying or hunting

–– 169 ––

activities.

Jonathan Haas (1990) described some settings in which archaeologists can

interpret plausible defensive purposes.

Nevertheless ... [concerns of warfare] may be objectified in the construction of defensive features, such as walls or moats, or in the deliberate selection of defensible site locations. “Deliberate” selection of such locations may be inferred when ready access to resources, water or arable land is sacrificed in exchange for elevation, difficult access, unrestricted or strategic vistas, or physical protection from attack (Haas 1990:177-178).

In the following discussion of defensive sites, the case for the defensiveness of

these features and practices will be related directly to analogous examples, where

possible, of practices documented in ethnographic or ethnohistoric records, indicating

their use as defensive features. In other cases, the defensive aspects will relate to traits

mentioned by Haas (1990).

Defensive Aspects of Residential Villages

The Coast Salish implemented a broad array of defensive practices, many of

which leave archaeological remains or imprints. Evidence of these defensive types can

also be found in ethnographies, oral histories, and ethnohistoric documents. I discuss

regional variation regarding how defensive sites are employed. Defensive practices

were not limited to defensive fortifications alone, but were also an integral aspect to

residential village construction, house arrangement, and even village settlement choice.

Plankhouses

Plankhouses were more than just shelter from the elements. This is particularly

true in comparison to mat lodges or lighter structures that were used prior to the shed-

roof house, which first appears during the Marpole Period. Plankhouses were fully

wooden enclosures (Figure 5). Even the doors, ethnographers have noted, were

designed with protection in mind. They were closed and locked with crosspieces at

–– 170 ––

Figure 5: Example of plankhouse as fully wooden enclosure; from a detail of a photograph taken at a potlatch in Songhees territory in 1874 (Attributed to Albert Maynard, RBCM PN6810).

night (Suttles 1951:259). Stern (1934:99) noted that, among the Lummi, “the entrances to

houses [were] made difficult to prevent surreptitious entrance.” Eells (1976:23) described

house doors as having a “circular aperture cut through the building.... [One at Sequim]

was three and a quarter by three-quarters feet.... It was closed by sliding other boards

over the aperture.” Collins (1974:62) also remarked that Upper Skagit doorways were

both low and raised:

It was a small round hole cut in a plank so that it was above ground level. Persons entering the house had to step over the bottom edge and also to stoop. The rationale for this was that if enemies entered, they would be awkwardly situated and could be easily dispatched from within.

Suttles (1991:219) described that the intention was in part to force entrants into a

“vulnerable” position. Similar entrances were used for stockade walls, according to

Grant (1857:301), which further indicates the defensive aspect of these narrow entrances.

Moreover, they often placed the door on the narrow side of a house, which would also

give those within more latitude in their ability to respond to invaders (Suttles 1951:259).

In some cases, doorways would have protective entranceways, as “a door often had

plank walls that extended into the house” (Suttles 1951:259).

Plankhouse roof-tops were also employed to advantage. The Songhees used the

–– 171 ––

rooftops as lookouts. During times of expected raids, “men maintained a nightly watch

on housetops” (Jenness 1934). Also, during battle, the Upper Skagit would stand upon

the roof-tops and throw down “lighted torches or bowls of heated pitch or seal oil on the

attackers” (Collins 1974:115). Inside the plankhouses, defensive features were also

constructed. Frank Allen of the Twana noted that a “trap door” was used for escape,

mentioning a story where a man exited his house through a tunnel after being was

awakened by his dog (Elmendorf 1993:129).45

As Duff (1952:47-48) described, there were two primary types of shed-roof

plankhouse construction, involving both detached, single-household entities and

extended, end-on-end shed-roof houses, all under one roof. In the sharing of walls, the

houses were more economical, requiring fewer planks to compose a household

compartment than a solitary plankhouse; however, this later shift also served to

aggregate the village population and minimize the avenues of attack for each household.

One such extended house, located in Suquamish territory, was called

Daxwklébeal, also called “Old Man House.” Gibbs (1877) described it as about 160 metres

(520 ft) long. Warren Snyder (1956) excavated at the house site and determined that at

least part of it was constructed by 1845, and noted that Chief Syáł or Seattle had lived

there. Hill-Tout (1904) similarly described one that was about 100 metres long in

Chehalis, while the earliest descriptions are provided by Simon Fraser in 1808, who

described one about 210 metres long in Matsqui. Fraser also described a similar

structure at Xwméthkwiyem or Musqueam (Schaepe et al. 2001). He described it thusly:

“The fort is 1500 feet [over 450 m] in length and 90 feet [nearly 30 m] in breadth” (Fraser

1960 [1808]:105-106). It is unclear from Fraser’s description whether he described a

stockade around a shed-roof plankhouse or that the extended plankhouse itself was fort-

like. Likely for similar reasons, Schaepe et al. (2001) put Fraser’s characterization of it as

45. As this account indicates, the common presence of dogs in villages also proved useful for warning of possible intruders.

–– 172 ––

a “fort” in quotes, as it was likely an extended plankhouse. If so, Fraser’s description is

suggestive of Suttles’ (1991:219) conception of the “house as fortress.” Suttles (1951:276)

acknowledged that these extended houses were “probably built for greater protection

from enemies.” Indeed, it has been noted that people often aggregate for protection,

forming either larger, concentrated villages, as in the Southwest (LeBlanc 2001; Haas

1990), or in this case, a larger house. Since walls separating households and families

were generally mat partitions, any intruders would meet greater numbers than they

themselves could bring in through a small doorway.

Household Arrangement

There are also sociopolitical aspects of defense, in which people are protected

unequally. Higher status individuals, for instance, generally selected the compartment

farthest from the door. Hill-Tout (1901, 1902, 1906; cited in Suttles 1991) repeatedly

stated that the chief occupied the safest place at the centre of the house, with commoners

and slaves near the doorways to take the brunt of an attack. Suttles (1991) commented

that this practice, where chiefs occupying the central compartment, was not typical. In

fact, with houses with one primary door, the safest location would have been the section

most distant from the door, not necessarily a central section. To approach a chief living

in a distant compartment, this meant that visitors would have had to walk farther to

approach the chief, in a way acknowledging the chief’s high status and serving to make

him less prone to attack. Slaves often slept next to the door (Schaepe et al. 2001:43):

The high-class smelá:lh (“worthy people”) occupied the warmest and safest portions of the house––most often the middle section farthest removed from the doorways and drafts.... The lower-status s'téxem (“worthless people”) and skw'iyéth (slaves) slept nearer the drafty doorways along the smokier back end of the house, serving as the early warning and defence system against intruders.

A similar principle operated for the village as a whole. Lower-class families,

among the Klallam, had to move outside the house with the slaves. As Gunther

(1927:183) detailed for the Klallam village of Suxtcikwí’iñ on Sequim Bay:

–– 173 ––

To the left of the trail from the bluff and at the beginning of the point of land in front of the lagoon stood a small group of huts in which the lower class of people lived.... The houses were small and poorly constructed. One of the pastimes of the young bloods of the upper class village was to come at night with poles and lift the roof from one of these small houses.

The placement in front of the upper class houses also made them more exposed:

Wherever a village had a lower class group they were always forced to occupy an open position, on a sandspit or an unprotected beach so that they would bear the brunt of an attack in war. While the enemy fought with this group, the people in the upper class village had time to prepare for the attack (Gunther 1927:183-184).

The distinction between the two classes within the above-mentioned Klallam

village, was marked by “a row of poles on which were put the enemies taken in war”

(Gunther 1927:184). Gunther (1927:183) also noted that the village had a stockade, but it

encircled only upper class houses.

Defensive Sites

According to ethnographic accounts, the Coast Salish employed an array of

defensive site types throughout their territory: lookouts, refuges, trench-embankment

fortifications, rock-walled defenses, and stockades. While trench-embankments were

palisaded or stockaded, I apply the term stockade primarily to palisaded fortifications

without trenches and embankments. Most of these defensive constructions were built

after 1600 BP and several types were documented during the postcontact period.

Lookouts

In many oral histories and ethnographies, lookouts are a key element in a

defensive plan. Lookouts occupy high-elevation spots with broad views of the coast or

passageways and several are noted throughout the Coast Salish area (Figure 6). The

name for one lookout on a mountain above the Fraser River in Stó:lō territory was

Alámex, which meant “babysit,” which suggests a metaphor for the time a scout would

spend watching from there (McHalsie 2001:141). Another lookout site in Howe Sound,

–– 174 ––

-126! -124! -122!

-120 

48!

50!

0 25 50 km

Washington, US-124  -122 

! !

! !

! !

Fort Nugent

Duckabush

Lookout

Maple Bay

Toba Inlet

Stitó:s

Pópkw’em

Qoqólaxel

Kay’kah-lah-kum

Desolation Sound

= Lookout w/ Rock Wall

= Lookout

! ! = Signal Station/Lookout

T atu’so

Defence Islands

Tsay-tsoh-sum (”facing outward”)

Alámex (”babysit”)

Lamalchi Bay

Figure 6: Lookout sites in the Coast Salish area as noted in ethnographic and archaeological sources.*

*A few rock-wall fortification sites are also in association with other defensive features as well, such as trench-embankment fortifications. Sources for these, and other defensive sites that follow, are listed in Appendix 1.

the Defence Islands are strategically located where Howe Sound begins to narrow as it

extends north-northeastward towards Squamish; it is associated with the name Tsay-

tsoh-sum, meaning “facing outward” (Reimer pers. comm. 2005; Bouchard, Miranda, and

Kennedy 1975:3).

Bryan provided one account of the use of a lookout, attributed to Chief Goliah of

Penn Cove on Whidbey Island.

–– 175 ––

The only information I received in answer to interrogations among the older white settlers of the survey area was from Mr. Ed Armstrong, who had questioned “Chief” Goliah about the entrenchment at Penn Cove Manor (Site IS-50). Goliah, host at the last potlatch given at this site, stated that when he was a young man some “northern Indians” were spotted in their canoes from the lookout station at Fort Nugent (Site IS-93). The lookout ran back to the village to give warning (Bryan 1963:76).

Lookouts have been described in the oral histories of the Klahoose, as occupying

major heights with views towards northern passages or on a bluff-top near the head of

Toba Inlet (Black, Urbanczyk, and Weinstein 2000).

A warrior of Duckabush Twana named Hwahwa’kwsǝb, built a lookout on a

promontory in Hood Canal about 1810. He “dug a hole down from the top of the bluff

on the north side of Duckabush. He dug it down and came out partway down on the

face of the bluff” (Elmendorf 1993:126). He then obscured the opening with brush.

Frank Allen stated that “hwahwa’kwsǝb goes up to his place every day to watch”

(Elmendorf 1993:126). One day, some Skagit raiders stopped on the beach below the

lookout, and the warrior killed every man in one canoe, while another canoe paddled

away. He enslaved four or five of the wounded (Elmendorf 1993:126-127).

After that [the warrior] had to keep a good watch, every day and every night, for the enemy to come again. After a while, canoes land right below his lookout. He hears their language; they talk like Skagit people. It was nighttime. Now [he] had lots of arrows his people had been making, had them up in his lookout. He didn’t say anything, just shot and shot down at the enemy on the beach. They couldn’t tell where the arrows were from. Lots of them were killed, lots wounded. They left one canoe there and paddled away with some of their dead (Elmendorf 1993:127).

Three canoes of Skagit men again returned to retaliate against him, but also

failed. His vigilance and the advantages of that lookout position were too great. From

such a position, only a bow or musket would be suitable. Notably, the attackers never

could place the origin of the arrows––an advantage of the bow over a musket, as the

thunder of a musket, while powerful with attendant psychological effect, does often

reveal the position of the sniper.

In Stō:ló territory, a lookout was constructed like a “watchtower,” located at

–– 176 ––

Stitó:s (McHalsie 2001:139, 149):

Most of the people lived at the Vedder Crossing (that’s where the highway bridge crosses the Vedder River.) Right at the point behind where the bridge is now there used to be a watchtower. They were expecting war canoes from other tribes up north, and they watched for them at this tower (Lerman 1952:145).

The informant, elder Bob Joe, told Lerman (1952:144) that he was describing

events of ”five to six hundred years ago." The Fraser River actually used to touch the

Chilliwack River at Vedder Crossing back then (Rafter 2000), so the watchtower would

have been well placed. In another mention of the Vedder Crossing watchtower, it was

noted that “In those days there weren’t any trees, and you could see from the Fraser

Valley to Vedder Crossing” (Lerman 1952:156). A specially constructed house was

located at the crossing as well, called Qoqólaxel, that had an inverted gable roof for

collecting water. This trough on the roof that “holds the water up” could be released to

make a loud warning signal (McHalsie 2001:137, 139; 145; Malloway pers. comm. 2008).

Another location in Stó:lō territory is at Pópkw’em, both a lookout and signal station. The

name is associated with “puff balls,” noted as likely associated with smoke signals

(McHalsie 2001:139, 144). Smoke signals as part of a lookout were also noted in

Snoqualmie territory, where foot runners would also dispatch (Tollefson 1996:155).

Another form of lookout has been found in the North Cascades Mountains.

Robert Mierendorf (1986; 2009 [pers. comm.]) has found two high-elevation pit sites

excavated into talus at locations offering lines-of-sight to Cascade Pass (45SK216 and

45CH754). The first of these (identified as FS 20; Mierendorf 1986) consists of eight pits;

the second site, recorded in 2008, consists of five pits. He interprets these talus pits as

lookouts monitoring travelers through the pass or as hunting blinds, or both. The

depressions would have provided some concealment and furnished some protection.

The talus pits are likely cultural in origin. They do not match mining or prospecting

features in style or setting; moreover, no historical materials are present in either site

while there are nearby precontact sites in the pass. Indeed, Cascade Pass has been

–– 177 ––

known as an important location for a long time.46

These types of sites also appear elsewhere on the Northwest Coast and Interior.

Prince (2004) determined that he had a lookout site associated with cache pits with

broad views of Kitwancool Lake in the Skeena River Valley, north of Fort Kitwanga

(GiTa-23). The site was 23 metres above the lake, and he acquired a radiocarbon date

from the lookout of 1300 ± 60 BP.

This site is atop a very steep, narrow ridge, barely wide enough to stand on.... This extreme topography was purposely altered at great effort to make it habitable. The crest of the ridge was terraced down to make a small platform, 5 m x 5.5 m, with a hearth in the center.... The position and limited size of this platform are more indicative of a lookout site. It has no easy route of access to the water’s edge below, but it has a 340 degree view-shed of the shoreline, including a clear view of the north part of the lake, and of the channel to the south, through which approaching canoes would have to pass (Prince 2004:49-50).

Lookouts were noted as one type of defensive site in the Aleutian Islands of

Alaska to the north (Maschner and Reedy-Maschner 1998). Also, in an overview of the

archaeology of warfare in North America, Lambert (2002) described lookouts as a site

type that should be apparent archaeologically. Lookouts have been a part of site

inventories from several archaeological surveys as well on the West Coast of Vancouver

Island. Brolly and Pegg (1998) noted several unrecorded lookout sites near Ucluth

Peninsula in Ucluelet Traditional Territory; Haggerty and Inglis (1984; 1985) recorded at

least six lookouts along Long Beach and the Broken Group Islands of Pacific Rim

National Park. Lookout sites were discussed by McMillan (1999:151-152), in

summarizing his own and other investigations in Nuu-chah-nulth and Ditidaht

territories. He also commented that promontory lookouts along the coast may also have

served as lookouts for whales.

46. Mierendorf (pers. comm. 2009; Mierendorf and Folt 2008) has also documented base camps inthe pass with a series of charcoal-rich hearths and pit features dating from 2010 BP to about 9500 cal. yrs. BP. The pass also has been known ethnohistorically and ethnographically as an important trade route. Collins (1974:13) also remarked upon the use of that route as one of the two main passes to the Interior from the Upper Skagit area. Boxberger (1996:49) described how an exploratory expedition in 1877 hired Upper Skagit individuals to lead them through the pass because of their familiarity with the route.

–– 178 ––

Archaeological manifestations of such lookouts might be light, although in some

cases, features may be present. Also, defensive sites in general contained lookout areas.

In Sechelt territory, the site of Kay'kah-lah-kum, a granite dome in Selma Park was a

defensive site with a lookout tree (Peterson 1990:28). Similarly, at the defensive site in

Desolation Sound, Menzies (1923:66) described an old maple tree used as a lookout.

Schaepe (2001; also 2006) described that as a primary aspect of rock-walled defensive

sites in the Fraser Canyon was to serve as lookouts. Indeed, the bluff-top settings of

many sites are situated to take advantage, not just of steep natural defenses, but also of

broad vantage points.

One such site was recorded in the North Cascades of Washington, Upper Skagit

territory. At a prominent overlook high above the confluence of Goodell Creek

(45WH490) with the Skagit River, a possible rock-wall lookout is present. The

archaeologist recording the site, noted its expansive views up Goodell Creek and the

Skagit River drainage. Between two large boulders, were two rock alignments built

with angular granitic cobbles and small boulders, one to the north, another with two

sections to the south (Kennedy 1992).47 The description appears similar to rock-wall

fortifications of the Fraser River.

Lookout sites are often located upon stony prominences with little stratigraphy,

except perhaps in niches and cracks. Lithic debris may still be scattered about but those

locations are also exposed to winds and storms that can leave them barren, if features

like rock alignments or niches are not present. One could surmise that this type of site

would be more difficult to locate archaeologically, although evidence of lithic debris,

and cache pits might be present. If inhabited for long periods, small hearth features

might be present, or perhaps debris from sharpening arrowheads or other weapons

might be found.48 The high and exposed location of these lookouts, especially ones with

47. At the time the site was first recorded, Mierendorf (pers. comm. 2008) thought that one wall was likely intended as a goat-hunting blind. The other did not appear useful as such, however, and he admitted that had they cleared the trees, it likely would have provided an broad vantage point.

48. In a discussion of inland shell middens, McLay (2004) noted that some were located on high –– 179 ––

rock surfaces, might lead to higher rates of erosion obscuring evidence of habitation, but

it is likely that some sites may still exhibit such features, and ground-truthing surveys

based on oral histories or sites identified through traditional use studies may turn up

such sites.

Refuges

The most basic form of seeking refuge is to send the women, children, and

elderly into the woods behind the village––to a place out of the way that would deter the

efforts of raiders. In the Fort Langley Journals, running and hiding was a common

reaction to news of Lekwiltok or Cowichan raiders en route upriver (Maclachlan 1998).

Refuges were a reasonable response to the threat of enslavement. Although men were

usually killed in warfare, women and children were often taken as slaves. If they were

away from the battle site, hiding in a refuge, the threat of enslavement was minimized.

The Upper Skagit, Snyder (1950-1954) noted, called these hiding places steetathl. They

referred to one such refuge as Sti el, located at the head of the Skagit River, consisting of

a camping area beneath a huge rock that had slid into the earth, creating a hidden space

beneath (Figure 7). In some cases, the areas were lakes located in the woods behind

villages. Two such sites were located in the forests east of villages on Puget Sound. One

of those was named seesáhLtub, or “calmed down a little,” which is likely associated with

its function as a refuge (Thrush 2008:220).

The Squamish had refuges in niches high on cliffs in Howe Sound (Reimer pers.

comm. 2008). Some Saanich groups living near present-day Sidney sought refuges

behind the village: “Behind them stretched a forest to which the inhabitants could flee

for refuge in case of attack” (Jenness 1934). When the British, in a gunboat, attacked the

village at Lamalchi Bay, they reported seeing women and children “carrying goods

away on their backs into the woods” (British Colonist 1863; cited in Arnett 1999). This

was a tactic employed even though there was a defensive structure, or “Block House” at

ground and might be the residue associated with “sentinel” or lookout positions.

–– 180 ––

-126! -124! -122!

-120 

48!

50!

0 25 50 km

Washington, US-124  -122 

Mt. Tzouhalem

Cave

Castle Peaks

Rockshelters/ledges

Sti el

= Rockshelter or Rock Ledge Refuge

SeesáhLtub "calmed down a little"

= Refuge Areas (Lake, Canyon, other)

Snoqualmie

Falls

= Blockhouse

Lamalchi Bay

Stillaguamish

“Stronghouse”

= Blockhouse

Figure 7: Locations of refuge areas and blockhouses.

the village. According to Stern (1934:99), the Lummi had refuges that were “prepared”

behind villages; it is unknown if he was referring to simply areas cleared and hidden

with stores of supplies or actual architectural features; in any case, these refuges behind

villages might be archaeologically detectable.

One type of unusual site may relate to these refuge areas. Inland shell middens

have been found located distantly from the coast and in some cases at high altitudes.

Over thirty of these sites have been found throughout the Gulf Islands, Vancouver

Island, and in the Sechelt area. McLay (1999) has looked at the diversity of shell

–– 181 ––

middens in the Gulf Islands. In a discussion of these rare inland sites, he noted that a

common hypothesis was that “these inland shell midden sites represent defensive sites,”

places of refuge deeper in the woods” (McLay 2004). Regarding several inland shell

middens adjacent to cliff escarpments on Gabriola Island, Wilson (1988:61) described

that they were “relatively small in area and are not particularly deep, suggesting

relatively short term occupation”; he also proposed a possible defensive origin.

Similarly, Brown (2000:43) remarked upon the unusual aspect of finding whole shells,

which would be much heavier for travel in comparison to dried clams; in his case the

site was seven kilometres from the coast. He found it “inconsistent with both

ethnographic and archaeological information on shellfish harvesting”; he posited that it

may been associated with periods of conflict. Although these may be associated with

other activities, defense remains a possibility and one could envision that the

archaeological signature of such refuge areas might look like an inland shell midden.

This tactic of sending women and children to refuges behind villages was

common enough that attackers used this to their advantage in some cases. According to

Stern (1934:103), the Cowichan, who were the “greatest foes of the Lummi,” attacked

one of their villages at Momli in Lummi Bay, just northwest of Bellingham. The

Cowichan sent many warriors behind the villages to capture the fleeing villagers.49 The

Lummi also employed this tactic. According to one tradition, a warrior named Skalaxt

wanted to avenge his brother’s murder by the Skalakin people. He spent years training

for this attack and told his assembled war party that there would be trails behind the

village and that he would hide along those when the rest of the party attacked the

village: “Give me a little start then follow up quickly and give the war cry....” As he

expected, he found the trail and waited the in the darkness and fog, mostly killing those

49. A plan that worked well, had not the Cowichans’ canoes lodged in the sand as the tide withdrew; the Lummi warriors simply caught the warriors at their canoes, regained their relatives, and nearly killed all of those Cowichan on the beach: “There were very few, some say two or three, who managed to get away during this battle by carrying a small canoe from the shore and using it to make their escape. When the tide came in, dog fish came and mutilated the bodies of the dead Cowichan. In referring to this incident the Lummi people say that the dog fish helped them in the battle” (Stern 1934:103).

–– 182 ––

who were headed along the trail (Stern 1934:119).50

There were strategies to deal with attackers along the trails outside their villages,

which indicates that this also must have been a common form of attack. As Sally Snyder

(cited in Bryan 1963) documented elders had known that trails outside villages

contained “camouflaged pits with upright spikes” to discourage or entrap such would-

be attackers.

With traps or not, trails often led to naturally defensive features, such as caves

high ledges. In other cases, the refuge area was modified to enhance protection, as the

Cowichan warrior, Tzouhalem, had done at a refuge cave (Jenness 1934). But Coast

Salish groups also constructed refuges.

Blockhouses

In addition to hidden refuges within the villages, some Coast Salish groups in the

southern Gulf Islands and northern Puget Sound built small “blockhouses” (see Figure

7, pg. 181).

Built of stout squared timbers, loopholed for muskets and cannon, the blockhouse allowed warriors to mass their firepower from a protected strategic strong point against an exposed, unprotected enemy. In the Saanich village of Tsouwat, each household built a blockhouse with plank walls (Arnett 1999:25).

At Lamalcha village on Kuper Island, there was one “Block House,” as the

British called it, in the center of the village. The British attacked the village because they

thought the murderer of a colonist was hiding out there, and “This ship was ordered up

there to teach those Indians a lesson,” according to elder Eddy Edwards (Arnett

50. “As they would come groping their way through the fog, he would club their heads and let them fall to the ground. The people seemed to be in a stupor and not notice what was happening to those ahead of them. When the bodies were heaped in a pile in one place, Skalaxt would move down the trail and continue to slaughter the people as they came. Each time he moved he would get a little closer to the houses until finally he came upon the warriors attempting to keep back their enemy while the rest of the people escaped. These men he also killed. The slaughter was so great that it is said very few, if any, escaped from that village.

“Skalaxt and his party returned victorious but he was not satisfied because he knew that the Skalakin tribe lived in many different villages” (Stern 1934:119).

–– 183 ––

0 100 200 m

KUPER ISLAND

= Plankhouse

= Blockhouse

Figure 8: Location and depiction of "Block House" at Lamalcha Bay village based on British accounts in April of 1863 (modified after Arnett 1999:137).

1999:131). Arnett (1999) described the account from the British Colonist (1863):

En route, the British “boarded several canoes and searched them” before arriving in Lamalcha Bay at 11:35 where the Forward anchored in front of the village. Seven or eight large houses stood along the beach inside a small crescent-shaped bay enclosed on both sides by thickly-forested points of land. At the centre of the village, in front of the lodges, was the Lamalcha fort, a blockhouse with walls eight feet high “strongly constructed of logs, properly morticed and loopholed on three sides for musketry.” Built with “squared timbers,” in imitation of hwunitum [settler] blockhouses on the San Juan Islands, the fortification was designed to defend the village against attack by musket-armed hwulmuhw [other Coast Salish]. Recently, it had been further modified with the addition of anti-artillery bunkers in the shape of “regular rifle-pits constructed inside the Block House, covered over with thick plank.” In addition, the blockhouse was protected “with numerous rifle-pits and trenches sunk around it.”

This “Block House” was much smaller than the numerous surrounding

plankhouses––less than half their size (Figure 8). While a refuge, it served a different

–– 184 ––

purpose than the underground house. The latter is used to hide women, children, and

the elderly, while in this instance, they simply resorted to hiding deeper the woods. The

blockhouse was used by the warriors who remained at the village. Moreover, its

location was front and center in the village. It is another instance of the Coast Salish

deploying multiple tactics in defense.

A similar structure was located at a site on a slough of the Stillaguamish River.

Like the blockhouse at Lamalchi Bay, it was within a village. Instead of planks, it also

was described as “built of big logs set on end, and a roof of heavy cedar slabs” (Bruseth

1972:11). This was called a “stronghouse” and it was the job of Tsalbilht, a warrior, to

maintain it. Villagers deposited their valued items and stores to be guarded by Tsalbilht

while they were gone on trips. There was also a trench dug around, although there was

no embankment, and the trench was hidden:

Around the house was a deep trench with a lot of sharp pointed stakes in the bottom. Over the trench was laid a network of sticks, and over this a layer of turf, with a secret firm path to the house. The idea was that enemy attackers or raiders would fall into the pit––and be impaled. It happened occasionally that Sklalams and King George Indians came in big raiding parties to capture slaves and valuables. Once a party of five strange Indians tried to rob the Stronghouse. Three fell in the pit, and two got away and went wailing down the river in the canoe (Bruseth 1972:12). Underground houses

Underground houses were used by the Coast Salish as a specific type of refuge

(Figure 9). Barnett (1944, 1955) first documented these semisubterranean features, also

called these “fighting houses[s].” Barnett (1955:49; 1944:267) described these features as

about “ten feet deep and rectangular,” with other informants describing these as “six

feet deep.” On top, there was a “flat roof of logs and planks laid flush across.”

Furthermore:

“Rafters” were laid at the pit edges to this variety of ridgepole. Poles, bark, and brush were placed over the rafters and the whole was covered with earth. There was no entrance from the top; a gangway sloped down to the floor level entry; and, for flight in case of attack, a tunnel led out the back way (Barnett 1944:266).

–– 185 ––

-126! -124! -122!

-120!

48!

50!

0 25 50 km

Washington, US-124! -122!

Penn Cove

Smelt Bay

Scuttle Bay

= Underground Refuge Sites

Skwokwílàlà

= Underground Refuge (General Area)

= Underground Refuge Sites (Archaeologically Documented)

Figure 9: Underground refuge locations in the Coast Salish area from archaeological and ethnographic sources.

Indeed, one of his informants described an underground house at Scuttle

Bay on the Sunshine Coast, as one of these refuges, which were occupied during

“troubled periods”:

The informant was certain that this was once a refuge, not in his father’s time, but perhaps during the lifetime of his grandfather; and furthermore, that it was not only a shelter for a day but was lived in over troubled periods. His father told him that the excavation was six feet deep with a flat roof of logs and planks laid flush across the opening and covered with earth. The entrance was near the corner and by an inclined approach. Curiously, the dirt walls were not planked up. Recesses were cut into them for beds and storage space (Barnett 1944:267).

–– 186 ––

Figure 10: Semisubterranean pit features at Penn Cove, Whidbey Island (45IS50), interpreted as underground refuges by Bryan (1963; detail from Figure 10).

McHalsie (2001:140) recorded the location of “pithouses” particularly built for

security in the Yale area of the Fraser Canyon. The name of the place, Skwokwílàlà,

translated as “hiding places.” Also, Chief David LaTasse of the Saanich noted that

underground refuges were built in preparation for Lekwiltok attacks:

We hear of them that they are coming, and we make ready. First we dig deep pits far up in the forest, deep pits with small openings but large as a lodge inside. That is to hide our women and children. We carry plenty clams and dried fish for them to eat, and all this we hide by bushes and trees, and cover our tracks so no man can find (Lugrin 1932).

Archaeologically, Bryan (1963; Figure 10) was likely the first to have proposed

such an interpretation, for “rectangular pits,” encountered on Whidbey Island. Sally

Snyder had described to him that the Skagit peoples had “hide-outs [that ] were

excavated at quite a distance in the back of a village for the women and children. These

–– 187 ––

pits were quite deep, and were covered by planks and underbrush” (Snyder pers.

comm. n.d.; cited in Bryan 1963:79). Bryan (1963:80) argued that the pit at Penn Cove

(45IS50) had “corner indentations on the lip of the rectangular pit” that he interpreted as

the inclined passageway. There is another semisubterranean pit at Penn Cove, which

likely represents another such feature, although given its prominent ramp, Bryan

(1963:80) called it a “horseshoe” shape, even though it is largely the same size as the

other pit. From his descriptions of the archaeological features and his informant

information, I believe Bryan is correct in his interpretations. His account may be the first

documentation of such a feature.

During my fieldwork at Smelt Bay (EaSd-2), I determined the presence of two

underground houses (UH 1 and UH 2). These meet Barnett’s (1944, 1955:54-56)

descriptions of such features in size and depth. Moreover, Barnett’s specific

identification of the “southwest corner of Cortes Island” indicates the site of Smelt Bay, a

place that also has been alluded to as having these features by other ethnographers

(Kennedy and Bouchard 1983:161). The traditional name for the site is Kw’úumáxen for

“shelter inside arm,” which Kennedy and Bouchard (1983:161) stated alluded to the

gravelly beaches that form a natural breakwater for the site; however, possibly it

contained additional meanings such as cultural protection––albeit a hidden one, just as

the houses were. A wireframe surface map indicates the deep excavation of the

semisubterranean pit (Figure 11). Notice the incline on its northern side towards the

adjacent plankhouse floor, which is also shown to indicate the surface level. Barnett

(194:268) noted that “Not every Muskwium family owned or had access to an

underground dwelling. Its construction was a family enterprise and was costly in

labor....” He also noted the Musqueam often used them for the sick or in times of

inclement weather, but that these “were decidedly a luxury” (Barnett 1944:268). Barnett

(1955:269) thought that these subterranean dwellings were like pithouses and were

adapted from Interior groups such as the Lillooet: “If we accept this, we must then be

prepared to admit that a dwelling has been modified into a refuge, and that the manner

–– 188 ––

Entrance Rampway

Area of lowest rim and gentler slope

Underground House

Floor excavated 1.5 minto older beach gravels

Adjacent Plankhouse

No rim of midden abutsunderground house to the southMidden Berms

scale in metres

Figure 11: Wireframe surface map of an underground house (UH 1) at Smelt Bay (EaSf-2) with plankhouse outline to the north (left) to indicate surface level.

of entry into an underground chamber can, and has, been altered from top to side in one

borrowing....” While Barnett noted some differences, his overall inference of Interior

influence appears to be incorrect. There are simply too many differences between the

two house types in form, function, and setting: (1) pithouses predominantly are circular,

while underground houses were rectangular; (2) pithouses exhibited conical roofs, while

underground houses had flat roofs; (3) pithouse entranceways were located commonly

at the top, instead of from the side (as well as hidden); (4) pithouses were primarily

residences, while underground houses served as temporary refuges; and (5) pithouses

were in prominent locations while underground houses were built behind residences or

villages, even sometimes located distantly from residential villages. However, like

pithouses, with grassy or other foliage growing on rooftops, one could imagine that by

obscuring the entrance, it would be well hidden.

–– 189 ––

Trench-Embankment Fortifications

Not always did they flee to the forests behind the houses. Occasionally they took refuge on rocky headlands impregnable on three sides, and protected on the fourth by a ditch and an artificial rampart of earth. ––Diamond Jenness (1934)

Compared to the previous forms discussed, trench-embankment fortifications

were of substantial architectural scale. Many were situated upon high bluffs with broad

views of the seascape. Steep bluffs, anywhere from 10 to 40 m high above the beaches

below, naturally formed a major part of the defensive structure, while a trench and

embankment were excavated along the exposed perimeter in flatter areas. Some took

advantage of ravines along either or both sides, heightening naturally steep defenses.

Others were situated on narrow and steep rocky headlands or high sandy peninsular

spits that afford the broadest possible view. These constructions required significant

investments of labour and likely were warranted only when warfare was commonplace.

Here I provide a description from Thacker (Smith 1907:385-386), who visited a one

trench-embankment fortification at Hunter Bay in the San Juan Islands:

On a recent visit to Lopez Island, I took the opportunity of briefly examining one of the ancient trenches, several of which are located there, and were apparently constructed for the purpose of fortifying certain points. The place I visited is situated on the southwest side of the island ... and consists of a bluff or headland several acres in area, jutting out somewhat into the water, with what appears to have once been a deep trench cut around its base on the land side. This trench commences on the west side of the bluff at the shore-line, on an almost perpendicular bank 7 metres or more above the water-line at high tide, and, running closely around the base for a distance of 100 metres, intersects the perpendicular wall of rock that forms the eastern side of the headland. The earth and rock thrown out were piled along on the outside of the ditch from the bluff, thus adding materially to its sheltering-capacity. The trench now varies from 0.6 of a metre to 1 metre in depth, and is about 2 metres across at the surface. At one place where the bed-rock comes to the surface, bowlders are laid along the line until the trench is resumed. At a point on the side of the bluff above the trench, and near where it intersects with the cliff on the east, a little nook makes back a short distance into the bluff, where the rocky background rises somewhat abruptly, forming a kind of miniature canyon, across the front of which appears to have been a wall of rock, which is now indicated by a line of small bowlders extending from side to side. This nook or corner would

–– 190 ––

Figure 12: Newcombe’s (n.d. [ca. 1935]) drawing of two trench-embankment types on southern Vancouver Island, including the peninsular bluff at Albert Head and a triple trench-embankment at Cadboro Bay.

accommodate a number of persons; and if protected by a covering overhead, such as an awning, they would be completely sheltered from the storms that frequently come in from the Straits of Fuca. This was the only sheltered place on the bluff. . . .For whatever purpose this trench was cut, it is run exactly where it should be for the purpose of fortifying the bluff by a rifle-pit, and I can conceive of no other purpose for which it could be used or constructed; and no better place could have been selected on that part of the island as a point of defence against a superior force, with so little labor, and at the same time hold so many advantages. The 100 metre trench connecting the perpendicular shore-line on the west with the rock-wall of the bluffs on the east, fortifies the land side, while its precipitous character fronting the water renders the place so nearly inaccessible that a few men could defend it against ten times their number.I found no evidence of burial inside the trench or in the fortified ground, nor any place indicating a water-supply, though it may have existed,––the one thing lacking to make this point an ideal fort, as the occupants could catch fish from the precipitous rocks on the water-front, and stand an almost unlimited siege, if they had water (Thacker in Smith 1907:385-386).

W. A. Newcombe (n.d.) first categorized fortifications across British Columbia in

the mid 1930s from trips taken with C. F. Newcombe and Harlan I. Smith (1927; 1934).

W. A. Newcombe described three primary types for the Northwest Coast, including

those in island, peninsular, and acclivity or bluff-top settings. He made depictions of the

two types found in the Coast Salish area (Figure 12): the bluff-like rocky headland of

–– 191 ––

Albert Head and the peninsular site at Cadboro Bay––island-style defensive sites were

more common for northern and West Coast groups.

Trench-embankment sites are the most broadly distributed type of defensive

structure built by the Coast Salish, although there is a concentration or core area near the

southern end of Vancouver Island (Figure 13). Judith Buxton (1969) provided a survey

of trench-embankment sites or “earthworks” in the Coast Salish area, many that are not

now in existence due to development. She classified these into three main types: Bluff,

Ravine, and Peninsular (Figure 14). Bluff-top defensive sites were placed high above the

coast, anywhere generally from 10 to 40 m above sea level. The main defense was

natural, consisting of the steep bluff which protects the front of the village. The setting

generally provided a broad vantage point upon which to view incoming raiders as well.

The edge of the bluff, however, only provided protection along the front, so in order to

protect the back of the village, a trench was constructed, with the same principle as that

of a moat in the Middle ages, although without water. The trench creates a steep

defense along the unprotected perimeter along the sides and the back of the site, a

cultural defense to complete the natural steep defense. Ravine defensive sites are

similarly placed atop bluffs, but exhibit deep ravines to either (or both) sides of the site,

generally gullies associated with intermittent creeks. These sites take advantage of

natural defensiveness even further, with the front protected by the bluff and the sides by

deep ravines. The defendants need only to create a shorter trench-embankment behind

the village, a slight arc to connect the ravines. These ravine types are the same (or a

subtype) as bluff type, and I prefer simply a bluff-top category.

Her third type consisted of a site generally closer to sea level, on a minor

peninsula or spit. The protected area is usually about 5 to 15 m above shoreline, but the

naturally steep protection nearly encompasses the perimeter of the site and, in most

cases, only a minor trench is needed across the neck of the peninsula. I have found that

these really involve two types that are situated upon quite different landforms: rocky

headlands as opposed to sandy peninsular spits. Rocky headlands are stony prom-

–– 192 ––

-126  -124  -122 

-120 

48 

50 

0 25 50 km

Washington, US-124  -122 

= Trench-Embankment Fortification

= Possible Trench-Embankment Fortification

Manor Point

Towner Bay

Cardale Point Indian Fort Site

Desolation Sound fort

Manson’s Landing

Rebecca Spit

Hunter BayAlbert Head

Cadboro Bay

Figure 13: Locations of trench-embankment fortifications.

Bluff PeninsularRavine

Figure 14: Buxton’s (1969:5; Figure 3) depiction of three major types of trench-embankments.

–– 193 ––

Bluff

Rocky Headland Peninsular Spit

Figure 15: Typology of trench-embankment sites for this investigation.

inences that are often peninsular, however, can be in locations as high as bluffs and

so rocky and exposed as to have little if any midden remaining. Whereas, those on

sandy peninsular spits generally were only 5 to 10 m high, and have broader expanses of

soil development for midden areas to accrete. Peninsular spits generally are in areas

with clam beds and other bay resources, whereas rocky headlands generally are distant

from such resources, surrounded by rocky shores.

For the purposes of the present study, trench-embankments are classified into a

new tripartite scheme by landform: Bluff-Top, Rocky Headland, and Peninsular Spit

(Figure 15). Archaeologically, these trench-embankment defensive sites have received

the most attention, and I will assess a couple of examples for each of these types of

fortifications. This will indicate the degree of similar principles behind the construction

of these sites as well as the variability of form as these were adapted to local settings.

–– 194 ––

Figure 16: Harlan I. Smith’s photograph of DgRr-5 from August 7, 1915, with a view of the trench,noted as “Looking east at earth wall from near south end” (Photograph 34035; 2 x 4 film from Geological Survey expedition, Canadian Museum of Civilization, Ottawa, Canada).

Bluff top fortificationsIndian Fort Site (DgRr-5)

On August 7, 1915, Harlan Smith visited a bluff-top trench-embankment in the

Lower Mainland of British Columbia, located along what is still called “Indian Fort

Drive,” one of which I provide (Figure 16). He wrote in his notes that: “There is an

earthwork about 1 mile south of the station at Crescent, B.C. There is some shellheap

material inside on north edge and in ridge at east and south. It should be restored by

filling in paths made by cattle, and saved in a Dominion or Provincial Park” (Smith n.d.

[ca. 1915]).51 Smith described the size of the embankments at the site to be 12 feet wide

at the “ditch top” and 8 feet wide at the “wall base.” In another report, Smith (n.d. [ca.

1915]) described the whole site as a “semi-circular embankment about 4 feet high by 8

51. Unfortunately, the site was not protected as a park. Due to house construction on the site since the 1960s, the trenches have been flattened in landscaping and are not apparent, although coring investigations indicated that intact midden areas are present in limited areas (Angelbeck 2006). As Buxton (1969) has noted the majority of these trench-embankment sites have been levelled or otherwise developed, and the pace has continued since her study.

–– 195 ––

Figure 17: Drawing and interpretation of Indian Fort Site, DgRr-5, by Don Welsh (used with permission).

feet wide with exterior ditch about 4 feet deep by 12 feet wide, defending the side of a

small area on top of a bluff overlooking the sea....” (Smith 1915; cited in Simonsen 1970).

Don Welsh has made an interpretive drawing of the site based on his descriptions as

well as ethnographic descriptions of other nearby sites, particularly from Suttles field

notes (Figure 17); his drawing provides a depiction of light structures as sometimes used

at seasonal camps. The bluff is about 40 m high overlooking Boundary and Mud Bays,

with clear views across to Point Roberts peninsula. On the beach below, there are

several petroglyphs and probable canoe runs.

This site is an example of a bluff-top trench-embankment, and the fort was

protected by two steep ravine gullies to its north and south. With ravine settings, there

is extent of natural protection, while also minimizing the labour necessary in the

excavation of trenches.

Cardale Point (DgRv-1)

Cardale Point is bluff-top trench-embankment on Valdes Island, on the first

triangular point north of Porlier Pass (Figure 18). In his study of Shingle Point, the next

spit on the same island to the north, Matson (2003:100) noted that the position of Cardale

–– 196 ––

Double trench-embankment Likely for double protection of entrance

Trench-embankment semirectangular,

extends from bluff edge to

bluff edge

Profile Trench 1

Steep Slopes/Bluff

scale in metres

Figure 18: Surface map of Cardale Point (DgRv-1).

Point would allow for ready control of Porlier Pass, one of the few passages through the

Gulf Islands between the mainland and Vancouver Island. There are two

portions of the site, the defensive portion up on the bluff and the older midden below

along the beach. have dated both parts of site, with the lower occupation dating back

over 4,000 years (4,130 ± 70 BP), while the fortification area dated to just over 500 years,

with three dates ranging from 510 to 540 BP (Grier and McLay 2001; Angelbeck 2009b).

A three-dimensional surface map of the site, produced with a total station, indicates its

position and the shape of the trenches (see Figure 18); two photographs of the trench-

embankment are also provided (Figures 19 and 20).

The site exhibits an oblique, subrectangular trench-embankment that protects

approximately 200 degrees or 55 percent of its perimeter, while the rest is along the bluff

edge, 15 to 20 m above the spit. In the southern portion of the site, the trench branches

in two sets of trenches about 20 m before the southern bluff edge. This might represent

–– 197 ––

Figure 19: View of eastern portion of trench embankment, with Eric McLay (top of trench) and Colin Grier (in trench).

a rebuilding and restaging of the trench, although since both maintain form, I argue that

it likely was a doubly protected entrance into the fort.

The trench that lines the back perimeter is also quite deep, taking advantage of

the natural prominence. The trench is so deep that it effectively serves as a double

protection––the outer embankment of nearly 2 m (at 45 degree slope) would have to be

breached, then a half-meter descent into the trench before a 55 to 60 degree slope up 2.5

m towards the top, where the base of the palisade wall would be located (Figure 21).

Core-sampling of the trench profile revealed, similar to other trench-

embankment investigations (e.g., Mitchell 1968; Buxton 1969), the slope was steepened

by the trench with the excavated matrix mounded in front of the trench (Figure 22). This

resulted in the removal of natural surface horizons in the trench area and natural

substratums overlying prior surface horizons in the embankment area. During our

–– 198 ––

Figure 20: View of embankment and trench from the east, Profile Trench 1 (Colin Grier and Eric McLay within fortified area on top).

–– 199 ––

10 11 12 13 14 15 16

-7

-6

-5

-4

-3

scale in metres

= Total Station Point

He

igh

t

Easting

Figure 21: Profile of eastern portion of trench embankment at Cardale Point as measured with a total station (Profile Trench 1).

11 11.5 12 12.5 13 13.5 14 14.5 15-7

-6.5

-6

-5.5

-5

-4.5

-4

EU 3

PR 7

PR 8 / EU 4

PR 9

PR 6

PR 5

PR 10

I

II

IIII

III/II

III

Estimated top boundary original surface (Zone II)

I

III

I

IIIII

I Dark greyish brown (10YR4/2) medium silt loam, topped by thin organic surface [ {O} A].

II Dark yellowish brown (10YR3/4) fine sandy loam [B1].

III Dark yellowish brown (10YR4/4) medium sandy loamwith small & medium gravels (15%) [B2 {C}].

III/II Mixture of Zones II & III

Symbol Zone Description

scale in metres

Figure 22: Stratigraphic profile of the lower portion of eastern trench-embankment (Trench Profile 1).

–– 200 ––

exposure of the trench and embankment, we also encountered a clear distinction

between the midden in the interior, which exhibited a hard demarcation that likely

indicated the placement of the palisade wall. It also exhibited a small postmould also at

the top of the trench (Angelbeck 2009b).

Peninsular spits

High sandy peninsular spits are also selected for trench-embankment sites.

Peninsular settings, again, generally provide the greatest amount of natural protection in

a perimeter, requiring only a single trench minimally across the neck of the peninsula.

At Cadboro Bay, however, three trenches were implemented, according to drawings of

Newcombe (n.d.; see Figure 12, see pg. 191) in the early 1900s. Sites on peninsular spits

are in locations that typically have other functions besides defense. Sand spits were

often near dense clam beds and likely were good areas for fishing. Hence the midden

surrounding or near the defensive site there could be quite deep, while the middens

within the actual protected area might still be shallow and spotty. I describe two

examples, both from the Northern Gulf Islands.

Rebecca Spit (EaSh-6)

Rebecca Spit is a defensive site situated at the front of a sand spit on Quadra

Island. The trench embankment is semirectangular, designed to steepen the slopes and

add obstacles along the southern front and western approach. Along the back to the

north, a longer trench extends nearly 50 metres across to the eastern slope, which is

naturally steep. Rebecca Spit is the most extensively excavated defensive site in the

Coast Salish region. Donald Mitchell (1968) conducted those excavations, including a

total of sixteen excavation units covering multiple aspects of the site, such as interior

midden areas, the fortified wall along the perimeter, and several profiles of the trench-

embankment feature. A surface map is provided of the site (Figure 23), which is

reconstructed based on the contour map provided by Mitchell (1968:30, Figure 1).

–– 201 ––

scale in metres

x = areas disturbed by road construction

Figure 23: Surface map of Rebecca Spit (EaSh-6), reconstructed from Mitchell’s (1968:30, Figure 1) contour map.

Mitchell noted that Heriot Bay was the largest village close to Rebecca Spit, located two

kilometres (1.3 mi) to the west of Rebecca Spit, while another semicircular trench-

embankment (EaSh-9) is located even closer, just 1.6 km (1 mi) to the south.

Within the upper, protected area of the site, Mitchell determined that there were

three small house platforms, suggesting less permanent structures than plankhouses at

residential villages. He interpreted these structures as “temporary,” however, there was

evidence that these were still “fairly substantial dwellings” (Mitchell 1968:44). This

indicates a lengthy occupation, if not a primary residence. Mitchell (1968:45) pointed

out that the absence of readily accessible fresh water would make it “untenable for great

lengths of time”; moreover, the midden areas within the walls of the site were “so

shallow that we are led to conclude the [site was] occupied for relatively short periods.”

Instead of a blanket of midden across the interior of the fortification, there was was a

scatter of shallow deposits with most near the “inner lip” of the perimeter. One

–– 202 ––

Figure 24: Mitchell’s (1968:32, Figure 4) stratigraphic profile of the top portion of the western trench-embankment feature, showing stakemould of barricade and midden material abruptly stopping at barricade wall.

hundred and twenty-seven artifacts were recovered. These were interpreted as a single

assemblage, and these included a chipped stone point, ground slate point, knife,

scrapers, and abrasive stones. Bone artifacts were more numerous, including 56 bone

points (or bone point fragments) of various styles (barbed, blunt-based, wedge-based,

and spindle-shaped bipoints). While it is known that many of these bone point types

can or did serve as points for subsistence––arming harpoons, fish hooks, leisters, or fish

rakes––most of these point styles can serve as arrowpoints. Mitchell (1968:37-38) noted

that the Comox Coast Salish, who had lived in the area when Rebecca Spit was

occupied,52 used several styles as points for arrows. Given the context of a fortification

site, it is more likely that many of these points served defensive rather than subsistence

function. Along the perimeter of the site’s high ground, Mitchell’s excavations indicated

the presence of stakemolds; he provided a profile of a postmould from the top of the

western trench-embankment (Figure 24). In some units, they recovered the remains of

cedar stakes. This and other stakes revealed in postmoulds were “clearly pointed” for

52. In the early 1800s, the Lekwiltok expanded southward taking some of Comox territory including Cape Mudge on southern Quadra Island (Taylor and Duff 1956).

–– 203 ––

insertion vertically into the ground (Mitchell 1968:40). This line of stakes marked a

“vertical break” of the midden area of gravel and shell inside the wall. The pattern of

stakemolds along the outer perimeter suggested to Mitchell (1968:33, 44) a “light

barricade.” Or, since these were so widely spaced, these may have indicated light posts

supporting a wall of horizontal cedar planks, similar to the construction of Coast Salish

plankhouse walls. Excavations of the trench-embankment feature indicated that

subsequent site formation processes have somewhat obscured the depth of these

trenches. For instance, the western portion was 70 cm deeper than the contemporary

surface indicated, resulting in a depth of 1.3 m behind the front embankment (Mitchell

1968:33). Making a case for defensiveness, Mitchell (1968:45) concluded “In each case

the ditches and walls serve to isolate a habitation area, and the most obvious explanation

for their presence is that their construction was primarily for protection.”

Manson’s Landing (EaSf-1)

At Manson’s Landing, there are two trenches (Figure 25). The first trench-

embankment was partially filled in by locals decades ago (Taylor pers. comm. 2007),

while the second trench-embankment still remains. The distance between the two

trenches is about 110 m. The inner area that is protected, between the second embank-

ment and the point of the spit, is actually quite small, only 38 by 23 m, or about 874 m2.

Rocky headlands

Trench-embankments constructed upon rocky headlands exhibit a similar

strategy as that employed at sandy peninsular spits in that only a narrow neck of land is

trenched. Otherwise, the landforms are quite different. While sand spits were often

near beaches and clam beds, rocky headlands are surrounded by cliffs or rocky

shorelines. In some cases, defensive sites on rocky headlands often exhibit minor

midden areas, simply because some have less areas of soil development.

–– 204 ––

MANSON’S

LAGOON

scale in metres

First Trench-Embankment across neck of spit (partially filled in)

Second Trench-Embankment

Protected Area

Figure 25: Surface map of Manson’s Landing trench-embankment site (EaSf-1).

EaSd-3

In the summer of 2005, I was able to conduct some limited investigations in

Desolation Sound (Angelbeck 2009a). This consisted of a rock bastion with two small

bays on either side. It is a natural fortress, yet there was much evidence of further

constructions to make it even more protected (Figure 26). From the beach and midden

area below, a narrow path led steeply upward to the flat on the promontory. The path

up appeared constructed as a narrow ramp such that only one person at a time could

ascend. Steep, straight-sided stone bluffs dropped eight to ten metres to the bay waters

on all other sides. The top of the promontory was flat and open in vegetation, with a

large old maple tree. In three core tests conducted on top of the site, midden deposits

were sporadic and were thicker near the edges. A trench-embankment had been noted

for the site, according to an early site record (Archaeology Branch 1977), however, it had

either been filled in, or the report writer was referring to earthen embankments

–– 205 ––

Bluffsdrop

to water

Narrow Rampfrom mainland to top of headland

Protected Area with patches of midden

DESOLATION

SOUND

Embankments Along Ramp

scale in metres

Figure 26: Surface map of fortification in Desolation Sound (EaSd-3), view from the east.

protecting the ramp to the the north.

The site closely matches the description by Menzies (1923 [1792]:66), who

recorded seeing an abandoned fortification in Desolation Sound on Vancouver’s

expedition in 1792; Vancouver (1984 [1792]:64) also wrote about the site but did not go

ashore. It matches the location, within Homfrey Channel, and the landform

description––it even has an old maple tree, which Menzies (1923 [1792]:66) described as

having a platform that was used as a lookout. Here is his description, in full:

At the farther end of these Islands we come to a small Cove in the bottom of which the picturesque ruins of a deserted Village placed on the summit of an elevated projecting Rock excited our curiosity and induced us to land close to it to view its structure. This Rock was inaccessible on every side except a narrow pass from the Land by means of steps that admitted only one person to ascend at a time and which seemed to be well guarded in case of an attack, for right over it a large Maple Tree diffused its spreading branches in such an advantageous manner as to afford an easy and ready access from the summit of the Rock to a conceald place amongst its branches, where a small party could watch unobservd and defend the Pass with great ease.

–– 206 ––

Figure 27: Village near Bute Inlet with house-floors at top of slope extending outward protectively; detail of a drawing sketched by a member of Vancouver’s crew in 1792 (Vancouver 1798).

We found the top of the Rock nearly level and wholly occupied with the skeletons of Houses––irregularly arrangd and very crouded; in some places the space was enlarged by strong scaffolds projecting over the Rock and supporting Houses apparently well securd––These also acted as a defence by increasing the natural strength of the place and rendering it still more secure and inaccessible (Menzies 1923:66).

A similar architectural style to the “scaffolds projecting over the Rock” was

employed near the mouth Bute Inlet at a likely Homalco village, according to a drawing

from the Vancouver Expedition (Figure 27). The drawing shows housefloors

overhanging the edges from the top of the slope, representing another instance where

plankhouses were situated in defensible locations, while not being fortified. Indeed,

Menzies’ description indicates that Site EaSd-3 was not palisaded, likely because the

sides were so steep.

Manor Point (DbRv-13)

Manor Point is located near the southernmost tip of Vancouver Island, on a stony

promontory facing eastward in the general area of Rocky Point; an area associated with

the highest concentration of rock cairns in the Coast Salish area (Mathews 2006). In fact,

–– 207 ––

Trenched Areaacross neck of headland

scale in metres

Protected Area

Mainland

Bluffs10 to 15 m

to wateraround

headland

Figure 28: Surface map of Manor Point (DbRv-13).

one rock cairn is located about 30 m to the west of the trenched area. Between the

promontory and the mainland where the cairn is located, the landform narrows at the

neck of the headland with steep, ten-metre drops to rocky shores along the north and

south; a surface map of the site is provided as well as a photograph of the trench from

the highest point on the bluff behind the trench (Figures 28 and 29).

The feature is distinguished from other sites of this type in that it is mostly a

trench, and does not exhibit an embankment in front of the trench. The trench, however,

is more substantial than most, with a depth of nearly a metre along the front, and

generally three to four metres wide. This accentuates the height of the rocky wall

behind the trench, which is over seven metres at the highest point from the top of the

wall to the base of the trench.

The main area behind the trench consists mostly of exposed bedrock, approx-

imately 65 by 40 m, with only spotty and shallow areas of soil situated primarily

–– 208 ––

Figure 29: View of trench at Manor Point (DbRv-13) from the east, from the highest point on the promontory behind the trench (Pete Dady in trench; Bill Angelbeck on western outer portion of trench feature; panoramic photograph by Darcy Mathews).

in niches between stony outcrops; these contained minor deposits with cultural material

such as lithic debitage. For this reason any structures within would likely have been

light and would have to be set up on top of the exposed rock. Areas of rock

outcroppings could have been used as naturally protective walls near the perimeter.

Rock-Wall Fortifications

Another type of defensive site has recently been identified in the Fraser Canyon

near Yale, consisting of rock-wall fortifications (Figure 30). Five have been documented

from Xelhalh, near Lady Franklin rock, to Lexwts’o’:kw’em, the narrows above Yale

(Schaepe 2000, 2001, 2006).53 These also make full use of natural defensive settings while

adding further rock-wall protections along points facing the river.

Most were located in narrow places in the canyon with turbulent currents as

“natural barriers” to upriver canoes; these can be from less than a metre to over two

metres high and mostly composed of flat or elongated local rocks (Figure 31). These

53. Kisha Supernant (2008a, Supernant and Schaepe 2008) has continued research on these rock-walls in the Fraser Valley. According to preliminary results, the number of rock-wall sites has increased substantially (Supernant pers. comm. 2008b).

–– 209 ––

-126! -124! -122!

-120!

48!

50!

0 25 50 km

Washington, US-124! -122!

Hunter Bay

Mt. Tzouhalem

Mary Hill

Lexwts’o’:kw’em

Xelhálh

Skagit River

= Possible Rock-Wall Fortification

= Rock-Wall Fortification

Figure 30: Map of rock-wall fortifications in the Coast Salish area from archaeological sources.

sites were also at high places with good vantage of the river and with long line-of-sight

communication with other fortification sites. Schaepe (2001:52; also 2006) hypothesized

that “these sites were strategically selected as a series of guard stations involved in a

coordinated and co-operative multi-village effort aimed primarily at regulating river

passage into and through the canyon.”

Schaepe (2006:671) argued that this represented a “defensive network” in the

canyon (Figure 32), coordinating the efforts among the individual sites, noting that it

challenged “the long-held belief that individual households were the traditional centers

–– 210 ––

Figure 31: View of rock-wall feature at Xelhálh in the Fraser Canyon (Photo: David Schaepe).

of economy, and by extension, of political authority among the Aboriginal peoples of the

Northwest Coast” (Schaepe 2006:671). Still, while it challenged traditional models of

Coast Salish sociopolitical organization, Schaepe did not advocate a form of centralized

authority. Rather, he advocated a “corporate family group” model of sociopolitical

organization in which these defensive sites within the canyon network would operate

and coordinate with, notably, “a minimum level of intercommunity governance”

(Schaepe 2006:671). What Schaepe described fits will with an anarchic, decentralized

network that allows both household autonomy and broader alliances of coordination.

These major rock-wall fortifications were restricted to the Fraser Canyon,

however, there were rock-wall structures used in other parts of Coast Salish territory.

For instance, Jenness (n.d.) described that the Cowichan warrior, Tzouhalem had

constructed a barrier in a cave on a mountain now named after him: “In Mt. Tzuhelem

above he had a cave barricaded with rocks in which he could take shelter.” This site may

–– 211 ––

Figure 32: Network of rock-wall fortifications in the Fraser Canyon (from Schaepe 2001:52).

have taken on a more natural look with the intention of concealment (as a refuge) rather

than defense, but perhaps served both needs.

In the North Cascades of Washington state, a site on a high bluff and stony

outcrop contains two “rock alignment features” consisting of boulders and angular

granitic cobbles “piled between larger boulders and bedrock outcroppings” (Kennedy

–– 212 ––

1992). The site is on a promontory that overlooks the Skagit River valley and the mouth

of Goodell Creek. While it may be a hunting blind for mountain goats (Mierendorf pers.

comm. 2008), these also appear to be well situated as lookouts. In any case, the rock

walls are constructed in a similar manner to those described by Schaepe, and are in a

location high above the creek confluence with views potentially of those approaching

upriver.

Similar rock walls have been documented by Darcy Mathews (2004) on Mary

Hill, near the southern tip of Vancouver Island. These are U-shaped blinds that occupy

rocky exposures. These are strategically placed for they would have advantages for

attacking those coming around the tip from the south. Both rock walls may have been

within line-of-sight of each other, dependent primarily on vegetation cover in the past.

While more testing would be needed to verify whether these date to the precontact

period, there is some similarity of construction and strategic placement.

Rock-wall alignments have also been implemented at some trench-embankment

sites, as Carlson (1954:120-121) described for a fortification site on Lopez Island:

Site 215, located on a high cliff above Hunter’s Bay on Lopez Island, consists of a trench 90 feet long, 12 feet wide, and 3.5 feet deep. A 1 foot lip is found on both sides of the trench. Eighty-four feet to the north of the trench is a wall of stone slabs, 51 feet long, and 1.5 feet high. Associated with these features are twenty-two cairns and a shell mound.

While some trench-embankment sites employ two or even three trenches as

protection, Hunter’s Bay included a low rock-wall that appears to have been used as an

outer defense, in the manner described by Schaepe (2006).

Stockades

Often, in the literature, stockades are treated as if they were the same as trench-

embankment fortifications. However, some stockades did not have trenches or

embankments associated with them, and they typically are located on different

landforms. Whereas trench-embankment sites were located upon bluffs or peninsulas,

stockades were usually located on beaches or river banks. More differences between

–– 213 ––

these two defensive site types in a later chapter are discussed in Chapter IX. Here, I

argue that stockades should be regarded as a distinct category of defensive architecture

employed by Coast Salish peoples. These are distributed throughout the Coast Salish

area, predominantly on or near the coasts (Figure 33).

In the mid-1800s, Grant (1857:301) saw several of these stockades during his

travels around Vancouver Island. He described their entrances as having “A few round

holes, or sometimes low oblong holes or apertures in the palisades, generally not above

three feet high....” In this respect the entrances to stockades were similar to those

already described for the shed-roof plankhouse.

Stern (1934:101-102) provided a detailed description of a fort in Lummi territory:

Long logs were grooved along one side and fitted over these wedged points, each top log forming a section which was braced inside by other logs. The sections were so arranged that the stockade was rectangular enclosing the entire village. This required many sections for the village had two large houses ten to twelve sections each, at right angles to each other. Tunnels with rocks over the top were dug at opposite corners of the stockade to points a short way out so that the entire stockade could be guarded by two men at these lookouts. There was a large pole in the center of the enclosure for hoisting a pitchwood torch to give light in case of a night attack. They perfected the light so that they could see a dog from a distance at night. A plank was planted along the trail to the spring water directly in back of the stockade with sharp bone spikes protruding to hamper the enemy during attacks and to catch anyone seeking to poison the water supply. In the daytime, the spikes were fixed so that the villagers would not be hurt, but every night they were set again. The stockade was built by a man named Sneqwaniq. When it was completed bullets were sent to the tribes of the north as a challenge for them to come to battle but those messages were never answered.

Suttles (1951:322) noted three such stockades built in the Puget Sound region.

The fort at Blaine, for instance, he described as:

... consist[ing] of a stockade around two plank houses, with tunnels leading from inside to loopholes in the bank in front of the stockade. Inside were two poles upon which baskets of flaming pitch were hoisted to light the surrounding area at night. Similar features were indicated for the Lummi and Samish forts. The Samish fort also had poisoned stakes set around it (Suttles 1951:322).

These stockades also had features, in some cases along the walls, for a lookout.

–– 214 ––

-126! -124! -122!

-120!

48!

50!

0 25 50 km

Washington, US

-124! -122!

Káxtyo

Quilceda Creek

Kay-kah-key-ahm

Dungeness Spit

Suxtcikwí’iñ

I-eh-nus

Shingle Point

KeekullukhunKullukhun

Blaine

S.báliuqw

Gooseberry Point

Guemes

Cadboro Bay

Swinomish Fort

Penn Cove

Hócbale

Squaw Bay

Khenipsen

Saanichton

Lil’p’las

Hebolb

Homulchesum

Q laxad

e

Kay’kah-lah-kum

Qiq laxad

e

= Stockade

Salmon Bay

TutúhLaqs

Tlkotas

Grandma’s Hump

Edison Creek

Sand Hill

Figure 33: Map of stockaded villages from ethnohistoric and ethnographic sources.

As Jenness (n.d.) noted, a Cowichan warrior “Tzoxwlets fortified his village at Kenipsim

with a palisade, and had a man on watch all the time.” In the northern Gulf of Georgia,

large rocks were stored near the top of the palisades for throwing down on attackers

(Kennedy and Bouchard 1983:69).

Labour Organization

One striking aspect of these fortification sites, particularly the trench-

embankment sites and the stockades, is the amount of labour required to build them.

–– 215 ––

Julius Charles told Suttles (1948[2]:83) that, while the man with the carpenter power,

Xwłe’yukw, led the construction, the “Whole tribe worked on fort.” In another

interview, Charles told Suttles (1949[5]:70) that the “Lummi didn’t go month [away].

Had to build forts to protect the people.” That is, the Lummi forewent other activities so

that they could invest time in building a fort for a month. During that time, they had to

rely on stores of food instead of building up their surpluses. Also, while Xwłe’yukw led

the people of his own village to build a fort, Charles said that the Lummi did not have a

man with such power and had to hire a Samish man, named Syǝqwa’nǝq, to lead the

construction (Suttles 1949[5]:70). So, in addition to investing labour for its construction,

the Lummi also had to hire a specialist to direct and plan the work.

As fortifications generally had room for a few households, the hiring and

construction likely would have been shared by the household chiefs, each expending

some capital for such investments. They might also earn some social and symbolic

capital through the organization of such efforts, just as elites might earn capital through

organizing household activities, as detailed by Grier (2001) for a Dionisio Point

household, and others (e.g., Arnold 1993). The construction of fortifications indicates a

further extension of controlling or organizing household labour. Indeed, the fear of

attack can be ideologically used to garner support for such efforts.

Ames et al. (1992) have provided some insight into the amount of labour

required for a single plankhouse for the Meier site in Oregon. They determined that one

house required about 40,000 board feet in building and maintenance throughout the

duration of its occupation, about four hundred years. Large numbers of planks and

posts were used in stockades as well. And, these had to surround not just one house but

multiple houses. For refuge sites, boards may have been borrowed from the main

village, as was done for some seasonally occupied villages, leaving mainly the

framework of posts––these skeletal houses gave early explorers the idea that these

villages were abandoned. From Suttles’ (1951) and Stern’s (1934) descriptions, the

palisade walls were constructed and do not appear to rely on planks from their houses

–– 216 ––

for temporary installation. For defense, it likely was more effective to have walls in

place and ready to protect as soon as needed, although interior structures might have

reused portable wall planks from their residential villages.

For most stockades, a palisade surrounded the full perimeter of a site, however,

Barnett (1955:38) reported that only the “most vulnerable sector” was stockaded. In his

excavations at Towner Bay, Mitchell (1968) found, parallel to the trench, a row of five

stakes eight cm in diameter that were placed high and inside the trench-embankment,

each of which was 25 cm apart. Again, at Rebecca Spit, Mitchell (1968:32) encountered a

row of stake remains at the top of the trench. And, as Kane (1971 [1847]) noted, when he

visited the fort at I-eh-nus, there were two walls: an inner wall that was only 5 feet high,

but the outer one with boards 20 feet high. I-eh-nus also was shown with planks, in

contrast to other descriptions of a wall of posts from young trees, which seems similar to

what Mitchell (1968) uncovered.

Among the sites in the northern Gulf of Georgia, surrounding Smelt Bay, the

inner protected areas of trench-embankment sites (meaning the area within the

innermost trench) average 48 m by 24 m, however, from the description of Snatelum

Point on Whidbey Island, the wall must have been at least 145 to 285 m long and 35 to 50

m wide to enclose the numerous plankhouses end on end within (Bryan 1963:47-48).

We must keep in mind that more than just stockade walls are involved in

construction. If a trench was present, a significant amount of earth movement was

conducted to create trenches commonly 2 m deep and 1 to 1.5 m wide and extending up

to 140 m in length as they protract in subrectangular fashion from bluff edge to bluff

edge, as at Cardale Point. Julius Charles described other constructions for one fort, as

Suttles (1949[5]:83) quickly recorded:

Fort––land sloping all around fence. 3 tiers of tunnels with loopholes. Inside [were] 2 houses [with] shed roofs plank walls. Fence has kind of sidewalk around with wall up to climb.

Tunnels for escape or entry require additional excavation and camouflage to

obscure. Entranceways would have needed boards for closing and locking, and

–– 217 ––

stockade walls require supports and cross-beams, and––as Charles described––a high

“sidewalk” for defense and a lookout. Lookout towers and areas would require

additional construction, as would torchlights. Then, additional efforts were also

required, including the making of weapons, assembling large rocks for tossing down

from the fort or gathering pitch for torches. Some also added carved elements for

intimidation or display of spirit powers, as at the fort at Lyack-son on Shingle Point that

Bishop Demers visited (Theodore 1939:187), and these carvings may have required

additional hiring of specialists.

What we have considered so far primarily regards construction––there were

additional preparations for battle. For such a stay, often away from fresh water or other

resource areas, supplies needed to be brought to the fort. For instance, in a

Snuneymuxw telling of the Battle at Maple Bay (Shtlup-netz), preparations started as

soon as the likely day of battle was known:

Five days before the time appointed for the battle, all the women and children are removed from the Nanaimo camp, and carried away to Chase River: where after laying up a plentiful supply of food, they are left to care for themselves, whilst the whole band of warriors, about six hundred in number, with thirty war canoes, started for the scene of the coming battle (Tate n.d.).

Conclusion

The Coast Salish employed numerous types of defensive constructions. These

included lookouts, signal stations, hidden refuges, underground houses, blockhouses,

trench-embankment fortifications, fortified rocky headlands, and stockades. Some of

these types were locally distinctive while others were shared among the Coast Salish

peoples.

Like much of the ethnography of the Coast Salish, the range of descriptions for

many of these defensive site types do not apply to the whole region. Barnett (1944,

1955;269-270) remarked on the underground refuges as primarily a northern trait, but

also practiced by Squamish and Musqueam, while others noted their presence in central

–– 218 ––

and southern Coast Salish areas (Bryan 1963:80; Lugrin 1932). Rock-wall fortifications

were used predominantly in the Fraser Canyon (Schaepe 2000, 2001, 2006), although I

have pointed to other limited examples in the North Cascades and San Juan Islands.

Trench-embankment sites have the broadest distribution, yet the greatest concentration

is along the coast of Southern Vancouver Island. Stockades and other fortifications were

described throughout much of Coast Salish territory, yet documentation of such

structures in southern Puget Sound is rare––and the Skokomish Twana were said to not

have built fortifications or have refuges, although they knew neighbouring groups did

(Elmendorf 1960:169).

While none of these defensive types can be said to be have been practiced by all

Coast Salish peoples, there are certain traits that were distinctive to the Coast Salish.

Whereas it is more common with non-Salish groups to the north and west to situate

their defensive sites on steep islets, which provided a full natural perimeter of defense,

such sites (and settings) are rare in the Coast Salish area. They preferred protected areas

on or connected to land––hence the need for trench-embankments. Rock-wall

fortifications may also be unique to the Coast Salish area, with such stone construction

not yet demonstrated elsewhere in the Northwest Coast, particularly for defense.

Furthermore, underground houses are yet another distinctive defensive practice among

the Coast Salish.

Mirroring their nature of anarchic social organization, the Coast Salish defensive

practices reveal a degree of local and regional autonomy and expression in the

architecture of defense. Yet, in their autonomy, these styles are not limited solely to one

subgroup of the Coast Salish, but rather reveal a sharing of practices across broader

areas, if not the Coast Salish area as a whole (Figure 34). This appears to match

distributions of unique stone bowls or burial mound constructions in earlier periods that

have been argued to indicate the sharing and alliances of an interaction sphere (Brown

1996; Grier 2003; Blake 2004). The pattern of distribution for these practices seems to

match the type of affinal alliance network described by Suttles (1987a [1960]), one that

–– 219 ––

-126! -124! -122!

-120!

48!

50!

0 25 50 km

Washington, US-124! -122!

= Core (Concentrated Area

for a Defensive Practice)

= Trench-Embankment Site

= Underground House

= Rock-Wall Fortification

Underground HouseCore Area

Trench-Embankment Core Area

Rock-Wall FortificationCore Area

= Peripheral Extent

(Links of Shared Practice from Core Area)

Figure 34: Core areas for three defensive practices and the peripheral extent for the sharing of each practice.

would have more nodes of alliance close by, but with certain individuals able to ally

with those farther away, and share their ideas and practices in turn. The underground

refuge represents a ready candidate for this. These were intentionally hidden,

sometimes even distantly from the residential village––awareness of these by potential

enemies would have critically hindered their effectiveness. Like a family’s bathing area

or ritual practice, the Coast Salish would have kept information about the sacred

location of a pool or the proper protocols for a ritual close at hand, maintaining its value

–– 220 ––

as cultural capital. Yet, with allies, knowledge and practices from a household are

shared with one another, just as food is shared and wives intermarried. And, this

practice––about a hidden feature––appears archaeologically in other regions outside the

northern Coast Salish, if less concentrated. This would appear to mirror the social

alliances that share these practices, with a core area in the north and a periphery where

some elites shared their practices with distant elites.

–– 221 ––

Chapter VIII: Defending Against Whom?

Interpreting the Purpose and Strategy of Defensive Sites

In the archaeological literature of the Coast Salish region, it is common to assume

that defensive sites were evidence of attacks by northern groups. In other words, it is

assumed that most conflict was intertribal, indicating it was between the Coast Salish

and non-Salishan groups. Coupland (1989:212) described trench-embankment

fortifications, for instance, as “probably defenses against northern raiders, who had

come south to trade at Fort Victoria, and intended to return home with Salish slaves”

(see also Ferguson 1984), although he also discussed a precontact possibility for these

fortification sites. Bryan (1963:76) also discussed “northern Indians” as a possible cause

for the building of these sites, and he provided an oral account of such an attack as an

example. Many rightfully have pointed (e.g., Mitchell 1968; Keddie 1996) to the advance

of the Lekwiltok or Southern Kwakwaka’wakw into Comox territory in the postcontact

period (Taylor and Duff 1956). Indeed, warfare did begin to proliferate after contact,

and many fortifications were constructed at that time. The Fort Langley journals of the

late 1820s, for example, reveal the presence and fear of northern attacks, as do many of

the oral histories (Maclachlan 1998). The concept that the forts were for protection from

non-Salish attackers is a recurring theme.

In this chapter, I evaluate these interpretations to see how they fit the data: not

just archaeological analyses, but also ethnohistoric, oral historical, and ethnographic.

First, I discuss the archaeological data, primarily looking at the distribution of defensive

sites in the Coast Salish area.

Although defensive sites have been known since the mid-1800s, archaeological

surveys and excavations were not conducted until the 1950s and ‘60s.54 Bryan (1955,

54. These investigations included Bryan (1955, 1963); Mitchell (1968); Buxton (1969), and –– 222 ––

GMT 2008 Nov 13 10:43:05 OMC - Martin Weinelt

-126 

-126 

-124 

-124 

-122 

-122 

47  47 

48  48 

49  49 

50  50 

51  51 

0 50 100 km

= Trench-Embankment Added by Buxton (1969)

= Trench-Enbankment by Bryan (1963)

-124 

49 

50 

0

-126 

48 

NUU-CHAH-NULTH

MAKAH BOUNDARY

KWAKWAKA’WAKW

LEKWILTOK BOUNDARY

= Approximate culture area boundaries ca. 1792

Figure 35: Trench-embankment sites recorded by Bryan (1963, Figure 12) and Buxton (1969, Figure 4) in relation to boundaries of Wakashan groups.

1963) presented a map of trench-embankment sites in 1955, highlighting those in

Northern Puget Sound––Bryan’s focal area for his survey––and other known examples

from predominantly Southern Vancouver Island. Fifteen years later, Buxton (1969)

conducted the largest study of trench-embankment sites (Figure 35). Many of the sites

she added were in the northern Gulf Islands, such as Rebecca Spit, previously

investigated by Mitchell (1968). She pointed to the paucity of defensive sites in the

Keddie’s investigations in the 1980s and ‘90s (Keddie 1983, 1984, 1987, 1996, 1996).

–– 223 ––

Central Coast Salish region: “Conspicuous by their absence are earthwork locations in

the Halkomelum region” (Buxton 1969:19).

From these assessments, as well as his own, Coupland (1989:212) interpreted that

“there is a tendency for some trench embankment sites to cluster near the Coast Salish/

Kwaguilth boundary, and near the Coast Salish/Nootka boundary.” Yet, even this map

contains several sites in the central region, particularly in the San Juan Islands and Puget

Sound. Furthermore, the southern Vancouver Island cluster is not so much at the Nuu-

chah-nulth boundary but rather more to the east, closer to the villages in the Victoria

region. In fact, one could argue instead these face the Olympic Peninsula and Klallam

territory to the south. Spanish expeditions did note such conflicts and tensions between

the Klallam and groups across the Juan de Fuca Strait (Wagner 1933:131), not tensions

with those on the coast.

I have drawn upon this earlier research and compiled and updated the data on

defensive sites, as discussed in the last chapter. The resulting map combines the array of

defensive site types (Figure 36). Some types of defensive sites did not leave an

archaeological signature, so it is necessary to combine information from multiple

sources––ethnography, ethnohistory, oral histories––to assess the range of Coast Salish

defensive practices. The map also includes sites in the central region––the Gulf Islands,

eastern Vancouver Island, and the mainland. In addition, it includes Schaepe’s (2000,

2001, 2006) recently documented rock-wall fortifications in the Fraser Canyon. The

relatively fewer sites recorded in the Halkomelem area may be due to the higher degree

of urban growth and development. Buxton (1969), in her survey, documented high rates

of destruction at those sites. In fact, their destruction was often a motivating reason for

investigations (Mitchell 1968; Keddie 1983, 1987, 1995).

The resulting combined defensive site map shows a more even distribution of

sites throughout Coast Salish territory instead of concentrations at the boundaries to the

west or north. This is not to say that the boundary areas were not zones of conflict––

those most certainly were. Rather, the concept of a borderline front of protection is

–– 224 ––

-126  -124  -122 

-120 

48 

50 

0 25 50 km

Washington, US-124  -122 

= Underground Refuge

= Blockhouse

= Lookout

= Signal Station

= Rock Wall Fortification

= Trench-Embankment Fortification

= Stockade

! !

= Rockshelter or Rock Ledge Refuge

= Lake Refuge Area

= Blockhouse

! !

! !

! !

Figure 36: Defensive sites in the Coast Salish area.

somewhat simplistic and, more importantly, does not accord with the nature of Coast

Salish sociopolitical organization. The idea of a borderline suggests a front line

protecting a centre, such as would be expected for chiefdoms or states. Coast Salish

sociopolitical organization had no such pattern of centralization. Defensive sites and

strategies were not more necessary at boundaries with neighbouring groups than they

were throughout the region. Defensive strategies were a necessary component of

political life for any village, or household, throughout the Coast Salish region. The

settlement pattern reflects this––in the map, the total array of defensive sites reveals a

–– 225 ––

widespread distribution throughout the Coast Salish area, with preferences for certain

types of features concentrated in different regions.

Conflicts with external groups were not restricted to or focused on border zones.

Lekwiltok raids, for example, were not concentrated on the southern boundary of their

territory. As was typical with most Northwest Coast warfare, attacks consisted of

surprise raids that could be directed at any village, including those deep within Coast

Salish territory. Communications networks would convey information about the raiders

and warnings (including false alarms) would quickly spread, often preceding their

arrival––a practice documented in the Fort Langley journals (Maclachlan 1998).

Moreover, attacks could come from any quarter, and not just Wakashan groups to the

north and west, but from Chilcotin (Kennedy and Bouchard 1983; Black, Urbanczyk, and

Weinstein 2000), Mid-Fraser groups (Duff 1952; Collins 1974), or the Chinook

(Elmendorf 1993); and later in the postcontact period, Haida and Tsimshian as well

(Walkem 1914; Curtis 1970 [1913]).

Coast Salish Sociopolitical Organization and Defense

Whereas subsistence and economy were organized at the household scale, the

organization of defense among the Coast Salish was conducted at the village scale,

involving significant cooperation among related households (Suttles 1951:277). This

indicates defense required a higher degree of communal interaction than was typical for

most day-to-day practices. Barnett (1944, 1955) found among the Musqueam, Sliammon,

Klahoose, and other groups, that preparations for warfare were still conducted at a

household scale, with some households owning their own underground house or

“fighting house.” Barnett’s description of household-organized defense suggests a more

autonomous organization.

Suttles (1951:277) described defense as being one of the few activities conducted

by the village as a whole, suggesting that households allied and cooperated with other

households in the village for defensive purposes, just as they would cooperate for large

–– 226 ––

potlatches in their village. As discussed above, authority might be granted to a

professional warrior in times of conflict, to whom all household chiefs would listen. His

authority only lasted as long as the threat, after which authority would then return to

the household chiefs.

Stockaded villages are an example of a village scale of defense. Also, some

trench-embankment fortifications are large, such as Cardale Point, where all the

households in the village below the bluff could occupy the fort; in fact, the fort area is

large enough that it may indicate the cooperation of other nearby villages, indicating a

scale of defense beyond the village.

Since the rock-wall fortifications in the Fraser Canyon exhibited lines-of-sight to

other fortifications and lookouts, Schaepe (2006) argued that these formed a network of

cooperation for defense. The linkages between these sites indicate a scale of cooperation

that goes even beyond what Suttles (1951:277) had described, and apparently in

contradiction to widespread notions of autonomous nature of households among the

Coast Salish.

However, this array of defensive practices––from household-scale to village-scale

defenses to inter-village networks––suggests a flexibility and variability that would be in

accord with the nature of defense needed for threats faced by Coast Salish. They did not

only face large-scale attacks from external groups. While much of the archaeological

discussion of defensive sites among the Coast Salish revolves around the attacks of

northern raiders or external groups, the ethnohistoric and ethnographic data as well as

the oral histories indicates more internal battles or feuds among Coast Salish groups

than with external groups.

Internal versus External Warfare

In three compendiums with accounts of Coast Salish warfare, the conflicts

described were often intercommunal. Various factors at Fort Langley recorded

–– 227 ––

0

4

8

12

16

Carlson (2001) Elmendorf (1993) Curtis (1913)

Types of Coast Salish Conflicts Recorded in the Colonial Period

External Conflicts

Internal Conflicts

(Versus Non-Coast Salish)

(Among Coast Salish)

Num

ber o

f Con

flict

Acc

ount

s

Figure 37: Types of conflicts in compendiums of accounts of warfare, ethnohistorically at Fort Langley (Carlson 2001) and ethnographically (Elmendorf 1993; Curtis 1970 [1913]).

observations on fur trading and native life from 1827 to 1830 (Maclachlan 1998). Carlson

(2001) found that during these years, over 30 conflicts were noted (Figure 37). Of these,

the majority (n=15) were between Coast Salish groups, the Cowichan raided up the

Fraser, the Klallam battled the Cowichan, the Snuneymuxw attacked the Chilliwack, and

so on. Conflicts involving non-Coast Salish groups, which consistently were by or

against the Lekwiltok (n=13), were also common, nearly even to the number of intra-

Coast Salish conflicts.

Another set of accounts are the oral histories of the Twana Narratives (Elmendorf

1993:126-164). Elmendorf was able to record fifteen stories relating to warfare and

comprising a significant portion of the overall accounts he acquired from the Allen

elders. Of these fifteen, ten accounts consisted of internal Coast Salish battles, while only

five related to conflicts with non-Salishan groups. Moreover, the non-Salishan groups

included a variety of opponents including Lekwiltok, Chemakum, Nuu-chah-nulth, and

Washington colonists. Curtis (1970 [1913]), in his volume on the Coast Salish, provided

thirteen accounts of conflicts, with the majority involving battles among Coast Salish

–– 228 ––

groups. These included Cowichan against the Sooke, Cowichan versus Klallam, and the

Klallam against the Sooke. Accounts about conflicts with external groups included the

Tsimshian, Chemakum, Lekwiltok, and Washington settlers.

Warfare with external groups likely were of a different nature, particularly

without ready protocols for conflict resolution. However, it seems apparent from these

compendiums that warfare was common between Coast Salish groups, even more so

than conflict with non-Salishan groups.

Occasionally, the Lekwiltok allied with bordering Comox household chiefs to

raid southern Coast Salish groups. Typically, this did not involve all household chiefs

from a single village, as some oral histories indicate (Duff n.d.; Tate n.d.; Cryer 2007

[1930]). Some Comox chiefs also acted in alliance with other Coast Salish groups against

the Lekwiltok (e.g., Jenness 1934). An example of the complexities of autonomy and

alliance among the Coast Salish is found in the accounts of the Battle at Maple Bay, the

final large battle with the Lekwiltok.

The Battle at Maple Bay

For many, many years, all in the bright summer weather, they have come down upon us, those Ukultahs of the North. They have killed our men and taken away our women to slavery. Every year they come, and nobody knows whose house shall be left desolate with the coming of the summer. For they are many and strong, and their war canoes are upon the sea as the salmon in the spawning season at the river mouth. We cannot stand against them. We are too few. We are not united as they are. Year after year we wail the loss of our champions, the loss of our wives and children.Then we make up our minds. All the tribes of the South, the Cowichans, the Malahats, the Songhees, the Saanich and the men from Sooke, where the tall white waves come in from the ocean––all of us make up our minds. We shall become one people and join and await the coming of the Ukultahs. They shall not find us until they come upon us all together.––Chief David LaTesse, Saanich (Lugrin 1932:38).

Having dwelt on the autonomous nature of defense and the frequency of intra-

Salish conflict, I would like to stress that there are also examples in the oral histories of

–– 229 ––

what can be considered moments of large-scale intertribal battle. The clearest example is

the Battle at Maple Bay, which likely occurred in the late 1830s (Angelbeck and McLay

2009). The accounts of this battle indicate many features of Coast Salish sociopolitical

organization, features that come into bold relief in the context of such a large threat. In

addition, the oral histories recount cultural practices and protocols that were followed in

preparation for the battle and its aftermath.

There are numerous accounts of the Battle at Maple Bay recorded by elders and

informants of the Cowichan, Snuneymuxw, Twana, Penelakut, Saanich, and Puget

Sound groups (Angelbeck and McLay 2009). In the full range of these accounts, there

was participation in the coalition from nearly forty Coast Salish groups, ranging from

Burrard, Capilano, and Musqueam in the east to Sooke and Songhees on the west;

Comox and Sechelt in the north to Duwamish and Puyallup in the south. Some accounts

are not as specific, mentioning only “Fraser River,” “Puget Sound,” or “Gulf of Georgia

Salish,” but even those categories in themselves suggest large groupings of Coast Salish

communities. Most accounts ascribe to the Cowichan a central role in calling the council

of war, where chiefs and warriors from various groups convened. Having just had one

of their villages devastated, the Lekwiltok claimed they would return to attack another

Cowichan village (Hill-Tout 1978 [1907]:160-162).

After this [a battle with the Lekwiltok] my people saw that something must be done. They had been nearly beaten that time, and their enemies were getting stronger. One day they would come down and finish the Cowichans. So they called the Indians to a big meeting at Lyack-sun, on Valdez Island. From Musqueam over to Esquimalt and Saanich, up the coast to Nanaimo, then down to Chemainus Bay and on to Valdez Island, they called all the fighters to come and talk about this thing, and see what could be done to stop those Indians from one day coming and beating them and taking all their women and children to be slaves.Well, the day came for that great meeting, and from all parts came the big canoes filled with the fighting men of the Cowichan tribe. The beach at Lyack-sun was filled with the canoes, and still there were more to come––the people from Musqueam and Esquimalt, and Saanich were not there yet (Cryer 2007 [ca. 1930]:141).

One remarkable aspect of these accounts is the repeated discussion of a “council

of war” that was called, nearly always in those terms––a typical procedure according to

–– 230 ––

Eells (1985:351). In most of the accounts, the Cowichan sent out messengers to other

Coast Salish villages throughout the Gulf of Georgia and Puget Sound announcing the

meeting. Most often, the accounts stated that council was held at Cowichan Bay, along

Cowichan River, or, in one case, at Lyack-Sun, or Shingle Point on Valdes Island.

Another account stated that the Snuneymuxw (Nanaimo) called the council of war,

although the meeting is in Shtlup-nets, or Maple Bay:

... Stah-qult, the old Nanaimo chief sent messages to all the Salish tribes along the southeastern coast of V. Isl., and across the Gulf of Georgia to the mouth of the Fraser River, calling their warriors to meet two days before the full moon at Shtlup-nets for a council of war. He has challenged the Laquiltoes (Tate n.d.).

According to Frank Allen of the Twana, there was a council of war for the Puget

Sound groups as well; perhaps this is a secondary council, as other chiefs may have

offered to gather others more distantly. At this meeting, which occurred on a long beach

on southern Whidbey Island across from present-day Port Townsend, the chiefs proceed

one by one to answer whether they would participate or not in this battle. First, each

made a public statement about their reasons for participating or not. Frank Allen,

whose great-uncle, “Big Jim,” had volunteered to participate in this battle, provided a

detailed account to Elmendorf (1993: 145-53):

All the Puget Sound war men met with the Nisqually to decide what to do. And they all asked one another, “What do you say now, Nisqually warrior man?’ What do you say, sαhe’wabš? What are you going to say, Snohomish? What do you say, Skagit men? What are you going to say, Swuqw’a’bš (Suquamish)? (Elmendorf 1993:145).

Much discussion followed among the chiefs and warriors, and attention

is given to the response of each group to the challenge of facing the Lekwiltok

(yǝkwiłtax):

“And then qaba’xad, the Snohomish warrior, said, “I’m going to die or kill yǝkwiłtax, one of the two. All the time they are raiding us Snohomish, and now I’m mad! I’m going to yǝkwiłtax!”“The big warrior from Skagit, dǝxwsdi’λαb said “I’m made to be a war man, and I’m not afraid of anything! I’m going to yǝkwiłtax!”“Now the Lummi speak, č’a’’wicut, the great Lummi warrior, got up and said, “The yǝkwiłtax have been troubling us too much! I might as well die

–– 231 ––

as not, so I’m going to kill them!”“And now kc’a’p, Kitsap, the big famous warrior of the Suquamish, said “I’m going to die or kill yǝkwiłtax” (Elmendorf 1993:145).

And so it continued, with Squaxon, Sawhemish, Skokomish, Gig Harbor,

Dungeness, and others. Not all present at the council joined in the alliance, however.

Leschi, a Nisqually chief, declined, although other Nisqually did join. A Skokomish

leader decided not to join, saying: “The yǝkwiłtax never bother me. And if they come to

my country, I’ve got warriors and I’ve got a trap to kill them!” (Elmendorf 1993:146).

There were many reasons for individual groups to decide to join the battle,

including events that occurred in the months before the battle. Boas (1889) described

how the Snuneymuxw and Sechelt carried out a retaliatory attack the Lekwiltok at

Qusan, or Salmon River, on northern Vancouver Island; Tate (n.d.) related the long story

of a Snuneymuxw chief’s son taken hostage and sold as a slave to northern groups in the

weeks before the battle. Hill-Tout (1978 [1907]:160-162) described how the Cowichan

villages were raided while most of the men were away––this was the final straw for

them, leading to the council of war. Two accounts also discuss a Lekwiltok council of

war and the performance of ritual warrior dances as well, as they were preparing to

mount a substantial raid on the Cowichan (Tate n.d.; Pearson 1969).

After the council of war was held, some immediately went into discussion of

preparations for battle. In some accounts, the preparations take up to a month

(Elmendorf 1993:146; Tate n.d.; Humphreys n.d.). Curtis (1970 [1913]:33) mentioned that

preparations occurred on Kuper Island while another account stated that the Puget

Sound groups prepared on Whidbey Island. There, they decided that Kitsap would lead

the Puget Sound groups in battle (Elmendorf 1993:146). They prepared their weapons,

canoes, and provisions of food. A shaman conjured spirit powers and several conducted

ritual preparations (Tate n.d.).

One account said that there were 200 canoes with ten men each, just from Puget

Sound, while Tate (n.d.) stated there were 4,000 fighters. Chief David LaTesse told how

they readied by sending the elderly, women, and children with stores of clams and dried

–– 232 ––

fish to underground refuges––as noted above, “as large as a lodge inside” (Lugrin 1932).

Tate (n.d.) described the arrival of 30 canoes of Snuneymuxw, 50 canoes of

Cowichan, 20 canoes of Chemainus, and 30 canoes of Musqueam and Tsawwassen,

totaling over 5,000 warriors. The numbers are likely exaggerated, but there is a

consistency throughout the accounts of an extraordinarily high number of participants,

especially when most Coast Salish endeavours typically involved only a few allied

households. Most expeditions, even for warfare, would have comprised of only a small

fraction of the warriors described in these accounts. No other accounts of warfare in the

region approach these numbers.

The Salish war leaders’ strategy was to confront the Lekwiltok on their terms.

They knew that if the Lekwiltok intended to go to the mouth of the Cowichan River,

they would have to pass through Samsun Narrows between Saltspring and Vancouver

Islands. Scouts were sent north of the Narrows to spot the advancing Lekwiltok on the

southward trek. Boas (1889:325) described how “Posts were continually maintained to

keep the tribes informed of the movements of the Lekwiltok and their allies.” According

to Arvid Charlie, a Cowichan elder, lookouts were also stationed along the approaches

to Maple Bay, both to the north and south (Angelbeck and McLay 2009). Eventually, the

Lekwiltok were scouted camped north of Maple Bay, and “’Hark! Hoo-ahoo-ahoole, the

enemy is coming’ rings out along the line of watchmen” (Tate n.d.).

In most accounts, the Lekwiltok were positioned north of Samsun Narrows. To

draw them through Samsun Narrows and into Maple Bay, the Coast Salish coalition

used a decoy: a canoe or set of canoes with women––a tempting prize of potential slaves

to ensure the Lekwiltok entered the bay (these were actually warriors dressed as

women).55 The Lekwiltok took the bait. Chief David LaTesse of Saanich described:

No need to paddle soft, think the Ukultahs [Lekwiltok]. The Southern Indians are afraid. They have fled before them. The noise of the

55. A couple of accounts have the Lekwiltok camped at Maple Bay and the Coast Salish using thedecoy canoe to draw them out onto the water, where they preferred to battle. For further discussions of minor variations, see Angelbeck and McLay (2008).

–– 233 ––

Ukultahs’ paddles against the sides of the canoes is like thunder, and they shout and laugh and sweep like a cloud into Maple Bay (Lugrin 1932).

Hill-Tout (1978 [1907]:161) recorded that the Salish groups had a set of calls to

organize when to strike:

A system of signals was also agreed upon. The sounds were to be those of the owl, the wolf, and the dog. The cry of the owl was to be given by the [Cowichan] as soon as they saw they were perceived by the Kwakiutl, the sound of the wolf when the Kwakiutl swallowed the bait and began to pursue them, and the sound of the dog would be given by those in ambush outside of the harbour to signify that they were ready to dash in and surround the enemy.

According to Humphreys (n.d.), a decoy is not mentioned, but he described that

the Lekwiltok entered the bay en route to the Cowichan River unaware of the warrior

groups hidden about the bay:

They passed on into Maple Bay, and headed towards the southern Narrows. When they got to about the middle of the bay, the division of war canoes that were left to guard the Southern Narrows came out to meet them. The Youlcatas stopped, and seemed to be considering. Then the party who were left to guard the northern Narrows launched their canoes, and blocked any chance of escape by that passage. When the Youlcatas found both passages disputed, they began to paddle towards the west side of the bay. It was then that the third division came out of hiding, and the Youlcatas found themselves surrounded by an enemy more than twice their own strength (Humphreys n.d.).

As Jenness (n.d.) put it, then “the Kwakiutl were threatened with a surround”

(Figure 38). Some accounts stated that, in seeing the Coast Salish alliance, the Lekwiltok

tried to be forestall a battle and began to negotiate for peace. In Frank Allen’s account,

the Lekwiltok “hoist up a white flag in their leading canoe. They don’t want to fight

now, they put up a white flag. But Kitsap puts up a black flag, that means he wants to

fight” (Elmendorf 1993:148).

A Cowichan warrior, according to Humphreys (n.d.), replied that “My people

will have peace when the Youlcatas are shorn of their power to fight”––then a warrior

fired an arrow into the Lekwiltok leader’s chest, who dropped into the water, initiating

the battle.

–– 234 ––

SCOUTS NORTHWARD

= Coast Salish Canoes

= Lekwiltok Canoes

= Coast Salish Stations / Roles (Scouts, Decoys, Snipers, etc)

SALTSPRING

ISLAND

VANCOUVER

ISLAND

X = Coast Salish Lookout

MAPLE BAY DECOY CANOE LEADING IN THE LEKWILTOK

SNIPERS / THROWERS

X

X

Figure 38: Map of the Battle of Maple Bay, indicating coordination of Coast Salish groups and roles.

–– 235 ––

During his telling, Chief David LaTesse “brought his hands together, fingers

sharply interlocked” (Lugrin 1932). As he described:

Like that the boats meet. We use our ... clubs made from elk bone. Thump. Thump. Thump. Down on the heads of our enemies. Every thump a kill. That noise––the rattle of bone upon bone––scream of the dying! But over all the triumphant war song of Thulpult and Quala Wonthult (Lugrin 1932).

In a Twana account, there is a crossfire of arrows:

And now the yǝkwiłtax come on, and when they get almost bow to bow with the Sound Indians they let go with their arrows, both sides shoot now as fast as they can pull their bows. The war man is in the bow of each canoe and the captain is in the stern, steering (Elmendorf 1993:149).

Frank Allen also described in particular how one Cowlitz warrior fired arrow

after arrow at the Lekwiltok:

Every time he shot, a yǝkwiłtax man would yell, “a . . . !,” and go overboard with an arrow in him! And some of those war men jumped right into the yǝkwiłtax canoes with spears or clubs, while the young men in their canoes went on shooting (Elmendorf 1993:149).

Other stories include descriptions of war powers being brought into play. Kitsap

told the warriors near him to not be scared of the lice that suddenly began to crawl

about since they were his power (Elmendorf 1993:148-149): “Don’t scatter my

ammunition now, just let my ammunition alone. That’s from my power now, just let

them crawl on you!” In another account by Angus August, a Cowichan, a man used his

rock power to lift up rocks near the surface as a reef to hold up or upend enemy canoes

(Bob 1980).

Nearly all warriors sang their power songs. The Cowichan sang their song of

Stimqua, the great warrior snake that long ago descended from the skies into Maple Bay,

linking the battle with the Lekwiltok to a mythic battle in the same locale (Harris 1901).

One warrior’s song made his enemies easier to defeat. According to Chief LaTesse: “It is

like a spell, that voice high and [ne]ar. To the dip of the paddle he sings, over and over,

the same song. All those Ukultahs must listen. They cannot help [it]. It is magic”

(Lugrin 1932).

–– 236 ––

The Lekwiltok fought back however. Curtis (1970 [1913]:34) described that the

Lekwiltok were “trusting to the greater size of their canoes to break through the

opposing line.” In a Twana account, Frank Allen described how spears were used to put

cracks in the Lekwiltok canoes:

Kitsap hollers to his captain, “Go right through between them now!” And his canoe goes right through between two of the yǝkwiłtax canoes. And Kitsap grabs up his short spear and stabs it into the yǝkwiłtax canoes, trying to split them. And the young men in his canoes do the shooting now, and Kitsap splits those two canoes and the water comes in and all the yǝkwiłtax go overboard (Elmendorf 1993:149).

This tactic was also described by Arvid Charlie, who related that these spears

were especially made for such use, possessing much larger spearheads than the ones

used for hunting. He also said that other warriors guarded the spearer’s flanks ensuring

that he was able to make the thrust. When the hull cracked, the enemy was concerned

with water flooding and sinking their canoes and so could not focus on fighting. Arvid

Charlie also noted that the smaller Coast Salish canoes, as mentioned above, were more

maneuverable than the larger Lekwiltok canoes, thus giving the Salish an advantage in

close quarters (Angelbeck and McLay 2009). Curtis (1970 [1913]:34) described how

Salish groups would also heavily lean their canoes as shields, “[throwing] their weight

to one side, raising the gunwale toward the enemy and depressing the other almost into

the water.”

In an account recounted by Ts’umsitum and Cryer (2007; also Cryer 1930), a

contingent of Lekwiltok canoes was driven by a line of Esquimalt and Saanich canoes,

pushing them towards bluffs within (or near) Samsun Narrows, where Coast Salish

fighters were laying in wait, prepared to attack. According to Louis Pelkey of East

Saanich, there were people with large rocks hidden and situated upon the bluffs on both

sides of the narrows (Suttles 1949[6]:51-54).

This was the chance the Cowichans up on the rocks had been waiting for. Just as the first canoe got under the bluff those men took great rocks that they had collected and rolled them down right into the canoe, breaking it into pieces. On came the next, and the next canoes! Too late now to stop, and no good trying to turn back, for our canoes were close behind them! Twenty canoes went under that bluff, and only three got

–– 237 ––

through (Ts’umsitum and Cryer 2007 [1930]).

In other accounts, such as Jenness’ (n.d.), Lekwiltok canoes faced rocks beneath

the surface of the water, snagging or wrecking upon rock reefs––again, by one account,

raised by one Salish warrior’s rock power (Bob 1980). As Curtis (1970 [1913]:34)

detailed, “some of them ran upon submerged rocks, and many were capsized. Those

that did not capsize faced a cordon of Coast Salish canoes, and some succeeded in

breaking through, but most of those that did so were overturned in the swirls of the

swiftly ebbing tide.” This made them ready targets for Coast Salish warriors. Either

swamped, split, sunk, or knocked overboard, many Lekwiltok “warriors were killed

with spears as they swam in the water” (Humphreys n.d.). In one account, they

described how the Lekwiltok were “speared like salmon” (Bazett 1910:6).

According to Florence James, some Lekwiltok canoes also fled to the shores of

Maple Bay (Angelbeck and McLay 2009). Curtis (1970 [1913]:34) noted, “Some were run

ashore, and their crews leaped out, only to be ruthlessly pursued and brought down.”

Not all suffering was by the Lekwiltok, as Puget Sound warriors were said to have died

and some canoes ran into trouble after exhausting their projectile ammunition––a

Lekwiltok arrow pierced one warrior’s eye (Curtis 1970 [1913]:16; Elmendorf 1993:150).

The battle, it was said, lasted without pause for a full day (Tate n.d.), two days

(Harris 1901), three days (McKelvie 1941), or four (Hill-Tout 1907). By its end, the bay

was “red with the blood of the slain” (Hill-Tout 1907; also Ronden 1913; McKelvie 1941;

Tate n.d.). Some accounts maintained that all the Lekwiltok warriors were killed, while

others tell of stragglers who made it ashore and hid in the woods––but eventually being

were killed when they exposed themselves searching for clams to eat along the shores

(Jenness n.d.; Curtis 1970 [1913]:34). Others were killed by Salish groups as they tried to

make it back north to home villages. The night the battle was over, the Coast Salish

groups built bonfires in celebration around Maple Bay (Tate n.d.), burning as fuel the

wreckage of Lekwiltok canoes (Suttles 1949[6]:53-54).

Curtis (1970 [1913]:34) mentioned that when he asked his Lekwiltok informants,

–– 238 ––

they “refuse to discuss this disastrous affair, frankly admitting, when pressed, that they

prefer to talk about their victories.” And, there had been many before. Indeed, in one

variant account56 by Curtis (1970 [1913]:14-16), even this battle is regarded by some

Puget Sound groups as an “ill-fated expedition,” as so many of their warriors had died.

This account is important as it highlights how the battle is portrayed somewhat

differently by different Coast Salish groups (see Angelbeck and McLay 2009).

Considering the number of different sources, the accounts of the Battle at Maple

Bay exhibit remarkable consistency in many respects. For a battle occurring in the early

to mid 1800s, the number of participants and the range of sources for the accounts, these

traditions exhibit common strains including: the council of war, the decoy canoe(s),

surrounding the Lekwiltok on open water, the rock reefs, rocks dropped from bluff

edges, and the redness of the bay. The battle is commonly described as the last battle

with the Lekwiltok, ending the major cycle of raids.

Moreover, the differences in these accounts stem more from which portions of

the battle they reveal: some discuss battle preparations, words or deeds just before

battle, or indicate aspects known about catching and killing stragglers. These kinds of

differences reveal distinct sources for the story, whether Cowichan, Saanich, or Twana.

These groups all participated in the battle, but each experienced it differently, likely

because each group participated in distinct theatres of battle, as attested in the more

detailed accounts (e.g., Humphreys n.d.; Tate n.d.; Hill-Tout 1907; Elmendorf 1993).

Differing histories of the battle are evident because each group highlights its own

own leaders and warriors rather than those of other groups. There was Tzouhalem for

the Cowichan, Dexwsdix'ab for the Skagit, Chidaskuid of the Puyallup, Ca'wicut of the

Lummi, Frank Allen’s great uncle Big Jim of the Twana, Kitsap for the Nisqually, Quala

of the Saanich, Hiloquib of the Duwamish, Stah-qult of the Snanaimuq and Thulpult of

56. Curtis’ (1970 [1913]:14-16) second account is considered a variant because it is about the Puget Sound groups against the Cowichan. However, it otherwise is remarkably similar, including the arrow in the eye of a Puget Sound warrior, as in Frank Allen’s telling (Elmendorf 1993:150); in fact, Elmendorf (1993:153) regarded Curtis’ second account as a variant of the Maple Bay battle.

–– 239 ––

the Cowichan, among many others. When taking these accounts in total, it cannot be

said which warriors were the most important, rather, many warriors were important––

and particularly important for the villages and groups they represented (Angelbeck and

McLay 2009). The variety of leaders and heroes, when viewed in total, appears as a

Coast Salish arena of contestation as each group proclaims its warriors’ lead role in the

decisive battle. That is, there is no hierarchy of leaders here, but a heterarchy of great

warriors. This extended to the council of war, where each group advanced its own

unique reasons for participating in the alliance. This was not complete Coast Salish

unity––after all, some declined to join the alliance and a few Comox were apparently

allied with the Lekwiltok. In the lead up to and during the battle, there was unity of the

Coast Salish, but once the external threat of the Lekwiltok had ended, these accounts

indicate contests over who was important and even about who was present and

participated in the battle. In other words, in true anarchic fashion, they returned to

more autonomous forms of interaction, as the need for such large-scale coalitions had

passed.57

After quoting Boas’ account of the first expedition against the Lekwiltok and

then Curtis’ recounting of the Battle at Maple Bay, Suttles (1954:46) remarked that: Evidently tribes from the Nanaimo to the Suquamish and the Skagit participated; the degree of cooperation and basis of organization, in what appears to be a rather loosely organized society, presents an interesting problem which has yet to be solved.

Suttles (1954:46) revealed this as a perplexing incident for a “loosely organized

society,” operating primarily at a household scale of organization. But had he viewed it

as an anarchic one, where networks of alliances can form bonds appropriate to the scale

and nature of the threat, such broad scale coalitions could be more readily

conceptualized. An anarchic form of organization retains its locus of autonomy at the

57. Some could argue that the gathering of Salish groups throughout Puget Sound for the signing of the Point Elliot Treaty in 1855 was another such moment, where the Coast Salish unified in great numbers, albeit not for war (Harmon 1998:226).

–– 240 ––

smallest unit––in the Coast Salish case, the household––but enacts and implements

alliances readily with other households and groups as it serves their interests, needs,

and values. As Carlson (2003:23) recognized, in discussing postcontact Puget Sound

groups, “The key” to understanding Coast Salish political affiliations was “to recognize

that they were built upon social networks which at certain times, and under certain

circumstances, could be operationalized into a formal political unity.”

An intriguing aspect of a Cowichan account of the battle is what happened

immediately before: Cowichan warriors had returned to their village to find it burned to

the ground with many slaughtered––the women and children were taken as slaves; as

Hill-Tout (1978 [1907]:160) put it: “Not even a dog remained.” Most of the Cowichan

warriors had been away, raiding and pillaging Coast Salish villages throughout Puget

Sound. It was then, on seeing this devastation upon their return, that they called on

other groups for a council of war. Among the many groups called to action were groups

in Puget Sound––where the Cowichan had just raided. In addition, many other groups

cooperated that had been recorded ethnographically to be “enemies.” In spite of these

previous enmities, these groups come together to avenge the Lekwiltok predations, both

recent and past.

The Contextual Nature of Enmity and Alliance

The oral histories about the battle illustrate a core principle of Coast Salish social

organization: groups act largely autonomously, but they organize together into larger

networks to meet specific needs or threats that cross-cut many groups. In the absence of

such conditions, local autonomy reigns. There is an Arabic saying that expresses this

concept: “It is me against my brothers; it is my brothers and me against our cousins; and

it is our cousins, my brothers and me against the world” (Barfield 2004:266).

To make this saying fit the Coast Salish situation, we should insert the household

as a unit between brothers (or immediate family) and cousins (extended family), and we

would need to add affinal allies created through marriage––but the principle is the same:

–– 241 ––

there is autonomy first, even conflict and tension, at the smallest scale (brother vs.

brother), but bottom-up unity to face larger threats. Moreover, it is a temporary unity or

alliance, lasting for as long as the conditions that generate it. Just as warriors were given

control of a village only for the duration of the battle, broader alliances were disbanded

at the close of the hostilities. This is just as Curtis (1970 [1913]:14) noted, “There was

constant internal strife ... among the Puget Sound Indians, but on rare occasions there

was cooperation for the purpose of checking the warlike northern tribes.”

These scales of alliance are depicted as increasing from the base in the household

and incrementally increasing in scale to include larger alliances with other households in

the village, regional networks, and distant affinal alliances (Figure 39). Each scale of

social organization has a corresponding material manifestation in an archaeologically

visible defensive feature: household scale of defense indicated by underground houses,

or even in Suttles’ (1991:219) description of the “house as fortress;” allied households or

villages cooperate in the construction of fortifications, both trench-embankment forts

and stockades; and defensive sites participate in regional networked defenses with

lookouts, signal stations, and lines-of-sight communication and access between forts.

Affinal alliances, the largest scale, are indicated in the overall distribution of defensive

sites throughout the Coast Salish area, representing a broad sharing of practices between

distant allies. Oral histories indicate such unity as demonstrated by the broad coalition

for the Battle at Maple Bay.

To sum up, while archaeological interpretations often emphasize intertribal

warfare, the ethnography, ethnohistory and oral histories suggest a great deal of warfare

among Coast Salish groups. The archaeological record shows––with the dense

distribution of defensive sites throughout the area, the pairing of defensive sites with

nearby villages and village-scale defense, household examples of defense, and networks

of defense––indicates multiple scales of defensive coordination. This material record

requires an interpretative framework that is consistent with the historical and

–– 242 ––

.

HOUS

EHOLD

VI L

LA

G

ER

EG

ION

AL

NET

WORKS

AFF

INA

L A

LLIA

NCESS

OC

IA

L O

RG

AN

IZAT

ION

FORTIFICATION

NETW

ORK O

F SIT

ES

B R O A D C

OA

LI T

I ON

S

DEFENS

IVE M

AN

IFES

TA

TIO

N

H O

US

E

UNDERGROU

ND

Figure 39: Scalar portrayal of social organization and corresponding defensive manifestation.

ethnographic records. Our interpretations must be consistent with our understanding of

Coast Salish social organization driven by the bottom-up.

–– 243 ––

Chapter IX: Autonomy and Alliance

Scales of Organization and Scales of Defense

In the prior chapters, I have considered the array of Coast Salish defensive sites

primarily in their form and spatial distributions. In this chapter, I expand upon the

network formation of defensive sites and examine temporal changes in defensive types.

In the last 1600 years there were two periods of warfare. The first occurred during the

Late Period, beginning with the transition and decline of the Marpole Period (with

defensive sites dating between 1600 to 500 BP) and the second, after Euroamerican

contact (ca 200 BP). What makes these periods distinct is the presence of

archaeologically visible defensive sites. In this chapter, I evaluate the scale of defensive

sites for both periods of warfare: both the scale of organization and the scale of

construction. The scale of construction for defensive sites is indicative of the scale of

sociopolitical organization. As discussed in the last chapter, Coast Salish political

organization allowed for high degrees of local autonomy at the household scale but yet

was ready to enable larger scales of social cooperation through networks of alliances, in

order to respond to military threats.

The complex sociopolitical organization of Northwest Coast cultures has long

confounded their classification within general anthropological models of sociopolitical

evolution. The quandary, as Matson and Coupland (1995:29) have formulated it, is that

the culture area “exhibit[s] high social complexity, but low political complexity.” For

this reason, the cultures of the Northwest Coast, and particularly the Coast Salish, have

been presented as exceptions to most evolutionary models of social organization in

anthropology (e.g., Fried 1967; Service 1975) . These models typically are constructed as

trajectories, often teleological ones, that lead to states; that is, these are models based on

centralization. As I have discussed throughout the previous chapters, Coast Salish

–– 244 ––

sociopolitical organization manifested in a decentralized manner, or in an anarchic

fashion. Here, I focus on how this anarchic sociopolitical organization is reflected in

Coast Salish defensive organization. I will begin with the Marpole/Late Period

transition, beginning ca. 1600 BP––a period when we see the initial expansion of

defensive sites.

The Distribution of Defensive Sites

Defensive sites, as discussed in the previous chapter, reflect an increasing scale of

organization, from household and village defenses to regional networks of sites. This is

a reflection of how households were politically autonomous units unto themselves. As

Suttles (1951:278) described, in discussing the establishment and construction of a

village at Saanichton, “In this case it is clear that the houses which made up the village

were built and owned separately.” That is, a village was not a cohesive unit but rather a

cluster of households.

Local autonomy of the Coast Salish has also been described by Marian Smith

(1940:6-7), who noted that categorical systems from other North American culture areas

“proved difficult” to apply to the Northwest Coast. She stated that “The organization of

such a people can best be described by stating the various affiliations to which men

might give their allegiance at different times and under different circumstances” (Smith

1940:6). She presented a model for scales of identification for southern Puget Sound. It

begins foremost with the family group, extends to the household, and then the village

group (notably not the village as a whole). Next, identification is associated with the

village in particular,58 and then the extended village drainage and broader watershed

58. Notably, the “village” may not refer to one specific location, but can sometimes can include several house clusters in proximity, as Collins (1974:15-20, 1980) also detailed for the Skagit River villages. Archaeologically, these would be recognized as separate sites, but these should be considered in a broader association. Snyder (n.d.) related to Bryan (1963:40-41) that two archaeological sites in Penn Cove appeared related as well: one site was for the “middle class” at Penn Cove Park (45SK50); across the cove, at Snatelum Point (45SK13) were the upper class houses. Snyder's informants stated that S'Golai-a and his family were able to move to Snatelum Point, likely indicating a promotion of status; this movement between classes is a topic for the next chapter, however, the point here is that individual and group identification described by Smith (1940) is not simply conceptual but manifests in spatial

–– 245 ––

system. It is a model of identification that is bottom-up: it begins with family first and

then involves larger scales of identification. However, these villages were not their year-

round occupation; while they had main villages, they also had seasonal camps and

fishing sites that would be lived at. A groups’ sense of identification to place would

have been tied to numerous locations.59

Suttles (1951:272) also noted that most social organization was organized on the

household scale:

The family was the basic unit in production and in consumption. It kept its own food supply and kept its own fire.... However, the more productive subsistence activities, the exchange of many kinds of possessions, the conducting of ceremonies, and defense from enemy attack required the cooperation of several families.

Suttles also emphasized that such cooperation was carried out through affinal

relations, marriages and alliances with other families. It is clear from Suttles that scales

of allegiance extend not rigidly up the increasing scale provided by Smith (1940:6-7), but

extend more branch-like through webs of allegiances even outside one’s own local area.

These connections are generally is not linked village to village, but rather the bonds are

household to household, just as marriages are arranged. There was also a preference to

marry or ally to a family of “a status at least as equal to its own” (Suttles 1951:289). This

is a dynamic for elites that in some ways encourages stronger ties to those distantly

located than with those lower class people from other households in the village. Over

time, without major changes or disruption, these repeated marriages and alliances with

others of similar status would perpetuate and accentuate class divisions.

The primary alliances between households and between communities

movement when that identification switches. 59. Florence James, a Penelakut elder, related this principle of identification to numerous places

in regards to her great-grandmother who lived at their main village at Lamalchi Bay. During a visit to the site, she described to us how they saw settlers moving in to their camps across the way on Saltspring Island. As she recounted, even though they were residing at Lamalchi Bay at the time, she said: “They lived there [on Saltspring Island]. That’s why it was such an offense. Because they didn’t just live here, they lived there, too” (Angelbeck and McLay 2008). That infringement led to the murder of one of those settlers, and the reprisal attack mentioned above (see pg. 183).

–– 246 ––

were those established by marriages. A marriage was ordinarily arranged by the families rather than by the couple to be married, and they arranged it with the benefits to be derived from the alliance clearly in mind. The wedding itself was the occasion for the display and transfer of privileges. Throughout its existence the marriage was the basis for an exchange of food and wealth between the two families. If the alliance was a satisfactory one it often continued to exist beyond the lifetime of one of the couple through the operation of the levirate or sororate (Suttles 1951:289).

Further, Suttles (1951:291:292) noted that the “more important” men had several

wives, reflecting their alliances with numerous households: Often the wives were from different communities; in this way a man established alliances with several other communities. The Lummi warrior sa’xwemgen had six or seven wives, one Klallum, one Duwamish, one Samish, one Skagit, one Lummi, one possibly Saanich, and perhaps another whose origin was forgotten (Suttles 1951:291-292).

Much of Suttles’ discussion of marriage alliances concerned the economic

advantages and privileges that a household might gain. But, the alliance also implies

defense. Family members were intertwined so that a threat to one household was also a

threat towards one’s family members in related households, following the principle of

social substitutability elaborated by Kelly (2000). Also, as these alliances allow

privileges––rights to access certain berry areas or clam gardens, for instance––a threat to

one household therefore becomes a threat to their allies’ privileges. Moreover, the

attacked household would in all likelihood call upon its allied households for food and

labour investment to rebuild. Affinal alliances confer on households the right to call

upon related households if their food stores were raided. This reciprocal relationship

ensures that related households assisted each others’ security. This mutual-interest

relationship meets what Tattersall (2006) described as a “deep coalition.” This is just to

point out that an attack on one household affects the potential or accessible capital of

their allies as well.

Suttles’ (1951:289) description of Coast Salish alliances is more detailed than

Smith's (1940:6-7). To be fair, Smith’s model is more about regional identification,

increasing from a household to the village to the villages that reside in a river valley, for

example. The model implies concentric rings of increasing social scale with the implicit –– 247 ––

notion that each larger unit of scale (household, drainage, or river valley) has a

sociopolitical coherence, when it often does not. Instead, those larger scales of unity are

dependent on allied households that may or may not constitute the whole drainage or

river valley at all. Suttles’ (1951:289-293) description of marriage alliances implies a

different sociopolitical dynamic: one that begins with the household and does not

extend simply to the next tier of neighbours up the drainage or inlet, but rather branches

out, near and far, to households in other villages. So, the pattern of alliances is in

practice rather more web-like with cross-cutting networks. Therefore, the increasing

scales of alliance would not necessarily extend out of the village and up the local

drainage area, but could extend in multiple directions to islands and other drainages

close and even quite distant. In fact, the greater the distance of an ally was usually

equated with greater prestige, a form of greater social capital. The more alliances, the

greater amount of potential organizational power (iii) one had and the more types of

resources (capital) a household would be able to access.

The bottom-up, decentralized sociopolitical organization of the Coast Salish

allowed households to be the predominant form of power, anchoring power in the local.

At the same time, there were principles that facilitated voluntary association with other

households in networks of alliances, after the form of increasing scalar organization

offered by Kropotkin (1910; e.g., see pg. 30 above). This dynamic of autonomy and

alliance manifests such patterns in the organization of defense as indicated in the

defensive sites through Coast Salish territory.

Archaeological Evidence for Networks of Defense

The increasing scale of defensive organization shifting from the household to the

region exhibits a networked pattern in the types of sites used for defense. First, as

discussed in the last chapter, there is a household scale of defense, or what was

described as the “house as fortress” (Suttles 1991:219). Also, there were external

household refuges employed by much of the Northern and some Central Coast Salish

–– 248 ––

groups. Barnett (1944, 1955) as discussed above, described these “underground houses,”

also called “fighting houses,” as hidden refuges, sometimes distant but also within the

village itself. His informants named four specific locales for these types of houses, and

one of which indicated Smelt Bay on Cortes Island. For the rest of this chapter, I will

focus on the village at Smelt Bay and related Late Period sites in the region as an

example of the increasing and networked scales of defense.

On the surface map of the Smelt Bay site, two prominent parallel ridges outline

the back for two rows of plankhouses and perpendicular ridges mark the plankhouse

side walls (Figure 40). Two of the house outlines, however, do not fit the typical

plankhouse pattern: both were not structures built on the ground surface but were

excavated up to 1.5 m into the surface, with the floors reaching into the old beach

gravels. These match the descriptions by Barnett (1944) regarding their depth and in

their rectangular shape. Moreover, only the two subterranean depressions are located at

the margins of the plankhouse site area. One underground house (UH 1) is located at

the southern extent of one row, while the other (UH 2) is located behind a house on the

second row.60 I interpret that the entrances as hidden ramps from adjacent houses, or as

he described them, a “gangway sloped down to the floor level entry” (Barnett 1955:49).61

These adjacent houses likely controlled access to these hideouts (See Figure 11, for a

detailed wireframe view of UH 1).

The ratio of regular plankhouses to underground houses is low, or about eleven

houses to two underground houses. Not every household had one of these. Barnett

(1944:268) indicated that subterranean houses were built by wealthy households who

could afford such investments in labour. In times of attack, households with

subterranean houses would be better protected. Contrary to Suttles’ (1951:277)

60. A possible third but much small refuge [UH 3] appears to be located behind another plankhouse to the immediate north of UH 2. It is very small in comparison to Barnett’s descriptions, but it may represent a refuge for hiding stores of food or other valuables.

61. Barnett (1944:268) also described these as “secret passageways leading from the plank house by a concealed opening in the floor.” He also mentioned “tunnels” leading out (Barnett 1955:49).

–– 249 ––

UH 1

UH 2

Plankhouse Outlines

(disturbance by

looters)

Area leveled for road

Possible small

refuge (UH 3)

behind plankhouse

scale in metres

Midden Berms

tideline

Figure 40: Surface map of Smelt Bay (EaSf-2), Cortes Island, indicating two underground houses.

suggestion that defense was one of the functions coordinated by the “village as a

whole,” these two subterranean houses indicate a form of defense coordinated by

households. Such an interpretation would be in keeping with the types of threats that

households faced, according to oral histories and ethnographies. For instance, attacks

were often directed not at whole villages, but aimed at particular houses, as would be

consistent with the nature of revenge attacks between feuding families.

Barnett (1955:268-269) noted that attacking parties scouted to “select the house or

houses to be attacked,” and “the party got away to their canoes before the village

defenders from other houses could counterattack.” Among the Coast Salish, given inter-

marriage ties, such assaults presented problems when the attacker’s were related to

some households in a village. For instance, Frank Allen described how Dungeness

Klallam warriors did not wish to attack all of the Hoodsport groups at a camp, so they

warned someone familiar with those at the site: “If there's anybody over there, in the

camp of the Hoodsport people, that is related to you, that you want to save, go and get –– 250 ––

GMT 2008 Nov 14 01:19:05 OMC - Martin Weinelt

-126!

-126!

-125!

-125!

-124!

-124!

50! 50!

51! 51!

0 5 10

= Large Village (Smelt Bay)

= Underground House

= Trench-Embankment Fortification

= Underground House (ethnographic)

= Probable Trench-Embankment Site

= Lookout (ethnographic)

Quadra Island

CortesIsland

MarinaIsland

HernandoIsland

DESOLATIONSOUND

STRAIT OFGEORGIA

Manson’s Landing

Smelt Bay

Rebecca Spit

0 2 4 kmBoulder Point

Cortes Bay

Gorge Harbour

Figure 41: Late Period defensive sites in the Smelt Bay region.

them ... when they come across the canal, we won't kill them” (Elmendorf 1993:131).

Also, one household may be in league with the attackers, as occurred in one Lummi

account, whereby “the members of one house ... by prearranged signal built a big fire in

their house and stayed safely inside while the enemy attacked the rest of the

community” (Suttles 1951:323). These examples indicate the autonomous actions of

households in both defense and attack. As discussed above, the murder of an individual

could readily escalate to retaliation by several of the victim’s household members. Since

there were attacks directed at particular houses, the coordination of a household

strategy for defense would appear to meet a common form of threat. However, since

only two households appear to have controlled these underground houses, there likely

would have been different defensive strategies for the other households, and there are

other defensive sites in the area surrounding Smelt Bay (Figure 41).

Several fortification sites are located nearby in a perimeter that radiates from

Smelt Bay, including other underground houses and several trench-embankment sites. –– 251 ––

According to Barnett (1944:267), underground houses were also present at the eastern

part of Cortes Island, likely the Cortes Bay area. Trench-embankment sites in the Smelt

Bay area include Manson’s Landing to the north (discussed above), two at the northern

and southern tips of Marina Island across the bay, and three on Hernando Island to the

southeast. Newcombe (n.d.) also noted a defensive site at Gorge Harbour; likely it is the

site located at the eastern side of the entrance (EaSg-6), which is located on a high bluff

at the entrance and exhibits a “series of dirt terraces....” that appear “man-made”

(Archaeology Branch 2008). Moreover, at Whaletown to the northwest, there is a spot

named for its association as a lookout;62 it has a broad view across the Strait of Georgia,

whereas the view northwestward from the village of Smelt Bay is largely obscured by

Marina Island. It has a view across the channel towards the fort site of Rebecca Spit.

While there are several trench-embankment fortifications near Smelt Bay, the size

of each is comparatively small to the size of the village at Smelt Bay, which exhibited a

minimum of thirteen houses extending approximately 260 m long and 60 m wide,

covering an area of 15,600 m2 (Table 3);63 there were likely even more houses but the

archaeological evidence for those has been obscured or destroyed by development in

recent decades, which means the overall size of Smelt Bay would have been even larger.

The trench-embankment sites in the area, however, are much smaller, averaging 47.8 x

24 m, or 1273.4 m2. It is unlikely that any particular fortification site was predominantly

for (or could adequately contain) all of the villagers at Smelt Bay, since the trench-

embankments are significantly smaller, less than 10% of the total area of Smelt Bay.

Since refuges are temporary occupations, a smaller area might have been tolerable for

62. There is oral history relating to the spotting of Haida warriors there, allowing for a warning before the coming attack (Black, Urbanczyk, and Weinstein 2000); afterwards, it was named T'ik'tn, or “place where you get discovered” (Kennedy and Bouchard 1983:155).

63. This area for the village of Smelt Bay is only restricted to the length and width of high midden berms that outline the houses; many of the berms and house outlines have been obscured by developments within the park and the adjacent private lots. The full midden extends over 800 m and is largely deep (often 1.5 to 2 m) and broad for much of its extent, so the area of 15,600 m2 is conservative.

–– 252 ––

Table 3: Size of Smelt Bay village, its defensive houses, and regional defensive sites.

Borden No. Site L W Area (m2) Average Areaby Type (m2)

RESIDENTIAL VILLAGE

EaSf-2 Smelt Bay* 260 60 15,600 15,600

UNDERGROUND "FIGHTING HOUSES”

EaSf-2 Smelt Bay UH 1*

10 25 250

EaSf-2 Smelt Bay UH 2*

12 13 156

Average 11 19 203.00

TRENCH-EMBANKMENT SITES

EaSf-1 Manson’s Landing* 38 23 874

EaSg-1 Marina Island S** 36 25 900

EaSg-2 Marina Island N** 25 13 325

DlSf-4 Boulder Pt, Hernando Isl.** 64 18 1152

DlSf-5 Hernando Island Southeast** 76 41 3116

Average 47.8 24 1273.40

*Measurements from Angelbeck 2009a**Measurements from Buxton 1969

limited duration. However, the size of these fortifications is more likely associated not

with the village as a whole, but with allied households. The underground houses

provide an example area that is suggestive for the area of a household scale of defense,

as UH 1 and UH 2 at Smelt Bay average 11 x 19 m, or 203 m2. The average area of the

nearby fortification sites, at 1273.4 m2, indicates three or four allied households, with

perhaps just two households at EaSg-1 on Marina Island (325 m2) to about four

households at Boulder Point on Hernando Island, DlSf-4. The largest trench-

embankment site in the area, DlSf-5 on Hernando Island, is only about one-fifth the area

of Smelt Bay.

There is a substantial disparity in the size of the Smelt Bay village to the average

size of defensive sites in the area (Figure 42). Two interpretations are possible to

account for this. Either a whole village could retreat to the confines of a much smaller

–– 253 ––

0

5,000

10,000

15,000

20,000

Smelt Bay Underground

Houses

Area of Residential Village Compared to Regional Defensive Site Averages

square

metr

es

Trench-Embankment

Fortifications

Figure 42: Smelt Bay village site compared to average size of regional defensive site areas.

refuge, and do so uncomfortably, probably tolerating such conditions only temporarily.

Some refuges are small enough, such as EaSg-1 on Marina Island at 325 m2, that the total

population of Smelt Bay are unlikely to have used it. Alternatively, only a few

households used those defensive sites. This latter interpretation, of a distributed

defense, is more in keeping with Coast Salish forms of social organization—rarely was

anything done on a village scale.

This might be an indication of a lack of village-scale defense, in contrast to

Suttles. However, considering Coast Salish traditions of distributed networks of

cooperation and alliance, the other sites in the area need to be considered for proper

context. Given the number of defensive sites in close proximity to Smelt Bay and their

Late Period contemporaneity, these smaller defensive sites more likely form a network